Files
Llama3-Chinese-pro-8.4B-sft-1M/trainer_state.json
ModelHub XC f437b0e601 初始化项目,由ModelHub XC社区提供模型
Model: linjh1118/Llama3-Chinese-pro-8.4B-sft-1M
Source: Original Platform
2026-06-29 16:06:13 +08:00

14022 lines
315 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9599692809830085,
"eval_steps": 500,
"global_step": 20000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 0.7233805656433105,
"learning_rate": 4.999999289433994e-05,
"loss": 1.595,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 0.6159764528274536,
"learning_rate": 4.999997157736379e-05,
"loss": 1.5717,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 0.6103418469429016,
"learning_rate": 4.9999936049083676e-05,
"loss": 1.4955,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 0.5304632186889648,
"learning_rate": 4.999988630951978e-05,
"loss": 1.5161,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 0.5091590285301208,
"learning_rate": 4.9999822358700397e-05,
"loss": 1.5431,
"step": 50
},
{
"epoch": 0.0,
"grad_norm": 0.5889601707458496,
"learning_rate": 4.999974419666187e-05,
"loss": 1.4517,
"step": 60
},
{
"epoch": 0.0,
"grad_norm": 0.5207070112228394,
"learning_rate": 4.9999651823448624e-05,
"loss": 1.5327,
"step": 70
},
{
"epoch": 0.0,
"grad_norm": 0.44091808795928955,
"learning_rate": 4.999954523911318e-05,
"loss": 1.518,
"step": 80
},
{
"epoch": 0.0,
"grad_norm": 0.5821202397346497,
"learning_rate": 4.9999424443716116e-05,
"loss": 1.4301,
"step": 90
},
{
"epoch": 0.0,
"grad_norm": 0.6290107369422913,
"learning_rate": 4.99992894373261e-05,
"loss": 1.4901,
"step": 100
},
{
"epoch": 0.01,
"grad_norm": 0.47097766399383545,
"learning_rate": 4.999914022001988e-05,
"loss": 1.5217,
"step": 110
},
{
"epoch": 0.01,
"grad_norm": 0.5277807116508484,
"learning_rate": 4.999897679188229e-05,
"loss": 1.481,
"step": 120
},
{
"epoch": 0.01,
"grad_norm": 0.6619521975517273,
"learning_rate": 4.999879915300621e-05,
"loss": 1.4222,
"step": 130
},
{
"epoch": 0.01,
"grad_norm": 0.49489903450012207,
"learning_rate": 4.9998607303492635e-05,
"loss": 1.4743,
"step": 140
},
{
"epoch": 0.01,
"grad_norm": 0.7624471187591553,
"learning_rate": 4.999840124345061e-05,
"loss": 1.6283,
"step": 150
},
{
"epoch": 0.01,
"grad_norm": 0.5447486639022827,
"learning_rate": 4.999818097299729e-05,
"loss": 1.4994,
"step": 160
},
{
"epoch": 0.01,
"grad_norm": 0.432623952627182,
"learning_rate": 4.999794649225786e-05,
"loss": 1.4901,
"step": 170
},
{
"epoch": 0.01,
"grad_norm": 0.6106896996498108,
"learning_rate": 4.9997697801365636e-05,
"loss": 1.4501,
"step": 180
},
{
"epoch": 0.01,
"grad_norm": 0.4500158429145813,
"learning_rate": 4.999743490046198e-05,
"loss": 1.5118,
"step": 190
},
{
"epoch": 0.01,
"grad_norm": 0.5799952149391174,
"learning_rate": 4.999715778969634e-05,
"loss": 1.5278,
"step": 200
},
{
"epoch": 0.01,
"grad_norm": 0.5897344946861267,
"learning_rate": 4.999686646922622e-05,
"loss": 1.4818,
"step": 210
},
{
"epoch": 0.01,
"grad_norm": 0.45949792861938477,
"learning_rate": 4.999656093921725e-05,
"loss": 1.4937,
"step": 220
},
{
"epoch": 0.01,
"grad_norm": 0.5662806630134583,
"learning_rate": 4.9996241199843105e-05,
"loss": 1.5197,
"step": 230
},
{
"epoch": 0.01,
"grad_norm": 0.4863593876361847,
"learning_rate": 4.999590725128553e-05,
"loss": 1.5607,
"step": 240
},
{
"epoch": 0.01,
"grad_norm": 0.5646289587020874,
"learning_rate": 4.9995559093734354e-05,
"loss": 1.552,
"step": 250
},
{
"epoch": 0.01,
"grad_norm": 0.5227863788604736,
"learning_rate": 4.999519672738752e-05,
"loss": 1.4672,
"step": 260
},
{
"epoch": 0.01,
"grad_norm": 0.45762133598327637,
"learning_rate": 4.9994820152450974e-05,
"loss": 1.3799,
"step": 270
},
{
"epoch": 0.01,
"grad_norm": 0.4994284510612488,
"learning_rate": 4.999442936913881e-05,
"loss": 1.4295,
"step": 280
},
{
"epoch": 0.01,
"grad_norm": 0.43154582381248474,
"learning_rate": 4.999402437767317e-05,
"loss": 1.4634,
"step": 290
},
{
"epoch": 0.01,
"grad_norm": 0.6734779477119446,
"learning_rate": 4.999360517828425e-05,
"loss": 1.4487,
"step": 300
},
{
"epoch": 0.01,
"grad_norm": 0.4380877614021301,
"learning_rate": 4.999317177121036e-05,
"loss": 1.5638,
"step": 310
},
{
"epoch": 0.02,
"grad_norm": 0.3950807750225067,
"learning_rate": 4.9992724156697875e-05,
"loss": 1.577,
"step": 320
},
{
"epoch": 0.02,
"grad_norm": 0.503062903881073,
"learning_rate": 4.999226233500124e-05,
"loss": 1.4237,
"step": 330
},
{
"epoch": 0.02,
"grad_norm": 0.4416813254356384,
"learning_rate": 4.999178630638297e-05,
"loss": 1.4839,
"step": 340
},
{
"epoch": 0.02,
"grad_norm": 0.5524111986160278,
"learning_rate": 4.999129607111368e-05,
"loss": 1.5262,
"step": 350
},
{
"epoch": 0.02,
"grad_norm": 0.4453815817832947,
"learning_rate": 4.999079162947203e-05,
"loss": 1.4764,
"step": 360
},
{
"epoch": 0.02,
"grad_norm": 0.3893362283706665,
"learning_rate": 4.999027298174479e-05,
"loss": 1.3778,
"step": 370
},
{
"epoch": 0.02,
"grad_norm": 0.4930141568183899,
"learning_rate": 4.9989740128226764e-05,
"loss": 1.3764,
"step": 380
},
{
"epoch": 0.02,
"grad_norm": 0.4468687176704407,
"learning_rate": 4.9989193069220874e-05,
"loss": 1.3774,
"step": 390
},
{
"epoch": 0.02,
"grad_norm": 0.4765455424785614,
"learning_rate": 4.9988631805038086e-05,
"loss": 1.399,
"step": 400
},
{
"epoch": 0.02,
"grad_norm": 0.5901889204978943,
"learning_rate": 4.998805633599745e-05,
"loss": 1.51,
"step": 410
},
{
"epoch": 0.02,
"grad_norm": 0.4665224254131317,
"learning_rate": 4.99874666624261e-05,
"loss": 1.4642,
"step": 420
},
{
"epoch": 0.02,
"grad_norm": 0.4224647283554077,
"learning_rate": 4.998686278465924e-05,
"loss": 1.4961,
"step": 430
},
{
"epoch": 0.02,
"grad_norm": 0.43521708250045776,
"learning_rate": 4.9986244703040144e-05,
"loss": 1.5362,
"step": 440
},
{
"epoch": 0.02,
"grad_norm": 0.5341386795043945,
"learning_rate": 4.9985612417920145e-05,
"loss": 1.4464,
"step": 450
},
{
"epoch": 0.02,
"grad_norm": 0.46132931113243103,
"learning_rate": 4.99849659296587e-05,
"loss": 1.47,
"step": 460
},
{
"epoch": 0.02,
"grad_norm": 0.41124364733695984,
"learning_rate": 4.998430523862328e-05,
"loss": 1.5451,
"step": 470
},
{
"epoch": 0.02,
"grad_norm": 0.4347441494464874,
"learning_rate": 4.998363034518947e-05,
"loss": 1.3869,
"step": 480
},
{
"epoch": 0.02,
"grad_norm": 0.5117555260658264,
"learning_rate": 4.9982941249740906e-05,
"loss": 1.3985,
"step": 490
},
{
"epoch": 0.02,
"grad_norm": 0.5037422776222229,
"learning_rate": 4.998223795266931e-05,
"loss": 1.4918,
"step": 500
},
{
"epoch": 0.02,
"grad_norm": 0.5024223327636719,
"learning_rate": 4.9981520454374484e-05,
"loss": 1.4403,
"step": 510
},
{
"epoch": 0.02,
"grad_norm": 0.5157750248908997,
"learning_rate": 4.9980788755264274e-05,
"loss": 1.4706,
"step": 520
},
{
"epoch": 0.03,
"grad_norm": 0.5240122079849243,
"learning_rate": 4.998004285575463e-05,
"loss": 1.5107,
"step": 530
},
{
"epoch": 0.03,
"grad_norm": 0.44245561957359314,
"learning_rate": 4.997928275626955e-05,
"loss": 1.4696,
"step": 540
},
{
"epoch": 0.03,
"grad_norm": 0.49761757254600525,
"learning_rate": 4.997850845724112e-05,
"loss": 1.4845,
"step": 550
},
{
"epoch": 0.03,
"grad_norm": 0.4365289807319641,
"learning_rate": 4.9977719959109494e-05,
"loss": 1.4089,
"step": 560
},
{
"epoch": 0.03,
"grad_norm": 0.480764776468277,
"learning_rate": 4.99769172623229e-05,
"loss": 1.4397,
"step": 570
},
{
"epoch": 0.03,
"grad_norm": 0.3849644362926483,
"learning_rate": 4.997610036733762e-05,
"loss": 1.4741,
"step": 580
},
{
"epoch": 0.03,
"grad_norm": 0.4486505687236786,
"learning_rate": 4.9975269274618033e-05,
"loss": 1.4565,
"step": 590
},
{
"epoch": 0.03,
"grad_norm": 0.448901891708374,
"learning_rate": 4.997442398463656e-05,
"loss": 1.4257,
"step": 600
},
{
"epoch": 0.03,
"grad_norm": 0.5016318559646606,
"learning_rate": 4.997356449787373e-05,
"loss": 1.448,
"step": 610
},
{
"epoch": 0.03,
"grad_norm": 0.37154853343963623,
"learning_rate": 4.9972690814818105e-05,
"loss": 1.4367,
"step": 620
},
{
"epoch": 0.03,
"grad_norm": 0.5966121554374695,
"learning_rate": 4.9971802935966345e-05,
"loss": 1.4547,
"step": 630
},
{
"epoch": 0.03,
"grad_norm": 0.447597473859787,
"learning_rate": 4.997090086182316e-05,
"loss": 1.4569,
"step": 640
},
{
"epoch": 0.03,
"grad_norm": 0.5071243643760681,
"learning_rate": 4.9969984592901325e-05,
"loss": 1.4707,
"step": 650
},
{
"epoch": 0.03,
"grad_norm": 0.47118982672691345,
"learning_rate": 4.996905412972172e-05,
"loss": 1.4747,
"step": 660
},
{
"epoch": 0.03,
"grad_norm": 0.3949645757675171,
"learning_rate": 4.996810947281325e-05,
"loss": 1.4194,
"step": 670
},
{
"epoch": 0.03,
"grad_norm": 0.37394997477531433,
"learning_rate": 4.996715062271291e-05,
"loss": 1.5161,
"step": 680
},
{
"epoch": 0.03,
"grad_norm": 0.6077260375022888,
"learning_rate": 4.996617757996577e-05,
"loss": 1.3857,
"step": 690
},
{
"epoch": 0.03,
"grad_norm": 0.5594815611839294,
"learning_rate": 4.996519034512496e-05,
"loss": 1.4215,
"step": 700
},
{
"epoch": 0.03,
"grad_norm": 0.4147029519081116,
"learning_rate": 4.9964188918751666e-05,
"loss": 1.3906,
"step": 710
},
{
"epoch": 0.03,
"grad_norm": 0.46012625098228455,
"learning_rate": 4.996317330141515e-05,
"loss": 1.407,
"step": 720
},
{
"epoch": 0.04,
"grad_norm": 0.40208500623703003,
"learning_rate": 4.996214349369275e-05,
"loss": 1.3953,
"step": 730
},
{
"epoch": 0.04,
"grad_norm": 0.4788183569908142,
"learning_rate": 4.996109949616987e-05,
"loss": 1.5743,
"step": 740
},
{
"epoch": 0.04,
"grad_norm": 0.3955346643924713,
"learning_rate": 4.996004130943996e-05,
"loss": 1.5613,
"step": 750
},
{
"epoch": 0.04,
"grad_norm": 0.498932808637619,
"learning_rate": 4.995896893410455e-05,
"loss": 1.47,
"step": 760
},
{
"epoch": 0.04,
"grad_norm": 0.4167149066925049,
"learning_rate": 4.995788237077326e-05,
"loss": 1.4475,
"step": 770
},
{
"epoch": 0.04,
"grad_norm": 0.4594554603099823,
"learning_rate": 4.995678162006371e-05,
"loss": 1.3917,
"step": 780
},
{
"epoch": 0.04,
"grad_norm": 0.5415817499160767,
"learning_rate": 4.995566668260164e-05,
"loss": 1.461,
"step": 790
},
{
"epoch": 0.04,
"grad_norm": 0.5038473606109619,
"learning_rate": 4.995453755902086e-05,
"loss": 1.4101,
"step": 800
},
{
"epoch": 0.04,
"grad_norm": 0.4519752562046051,
"learning_rate": 4.995339424996319e-05,
"loss": 1.4642,
"step": 810
},
{
"epoch": 0.04,
"grad_norm": 0.457968533039093,
"learning_rate": 4.995223675607858e-05,
"loss": 1.4132,
"step": 820
},
{
"epoch": 0.04,
"grad_norm": 0.4938353896141052,
"learning_rate": 4.995106507802498e-05,
"loss": 1.3892,
"step": 830
},
{
"epoch": 0.04,
"grad_norm": 0.40524962544441223,
"learning_rate": 4.994987921646846e-05,
"loss": 1.386,
"step": 840
},
{
"epoch": 0.04,
"grad_norm": 0.5541794896125793,
"learning_rate": 4.994867917208311e-05,
"loss": 1.4325,
"step": 850
},
{
"epoch": 0.04,
"grad_norm": 0.4559367597103119,
"learning_rate": 4.99474649455511e-05,
"loss": 1.4621,
"step": 860
},
{
"epoch": 0.04,
"grad_norm": 0.41948121786117554,
"learning_rate": 4.9946236537562675e-05,
"loss": 1.41,
"step": 870
},
{
"epoch": 0.04,
"grad_norm": 0.38462600111961365,
"learning_rate": 4.994499394881611e-05,
"loss": 1.4958,
"step": 880
},
{
"epoch": 0.04,
"grad_norm": 0.3599010407924652,
"learning_rate": 4.994373718001776e-05,
"loss": 1.4918,
"step": 890
},
{
"epoch": 0.04,
"grad_norm": 0.3832583725452423,
"learning_rate": 4.9942466231882044e-05,
"loss": 1.4486,
"step": 900
},
{
"epoch": 0.04,
"grad_norm": 0.3985520899295807,
"learning_rate": 4.9941181105131436e-05,
"loss": 1.5093,
"step": 910
},
{
"epoch": 0.04,
"grad_norm": 0.41597920656204224,
"learning_rate": 4.993988180049647e-05,
"loss": 1.4422,
"step": 920
},
{
"epoch": 0.04,
"grad_norm": 0.3992542624473572,
"learning_rate": 4.9938568318715725e-05,
"loss": 1.4571,
"step": 930
},
{
"epoch": 0.05,
"grad_norm": 0.3928009569644928,
"learning_rate": 4.9937240660535885e-05,
"loss": 1.4307,
"step": 940
},
{
"epoch": 0.05,
"grad_norm": 0.463318407535553,
"learning_rate": 4.9935898826711636e-05,
"loss": 1.3857,
"step": 950
},
{
"epoch": 0.05,
"grad_norm": 0.41009557247161865,
"learning_rate": 4.993454281800575e-05,
"loss": 1.4181,
"step": 960
},
{
"epoch": 0.05,
"grad_norm": 0.4060661792755127,
"learning_rate": 4.993317263518906e-05,
"loss": 1.4331,
"step": 970
},
{
"epoch": 0.05,
"grad_norm": 0.3853103220462799,
"learning_rate": 4.9931788279040445e-05,
"loss": 1.4223,
"step": 980
},
{
"epoch": 0.05,
"grad_norm": 0.413391649723053,
"learning_rate": 4.993038975034685e-05,
"loss": 1.4984,
"step": 990
},
{
"epoch": 0.05,
"grad_norm": 0.3775191009044647,
"learning_rate": 4.992897704990327e-05,
"loss": 1.4516,
"step": 1000
},
{
"epoch": 0.05,
"grad_norm": 0.4302771985530853,
"learning_rate": 4.992755017851276e-05,
"loss": 1.4718,
"step": 1010
},
{
"epoch": 0.05,
"grad_norm": 0.42789632081985474,
"learning_rate": 4.992610913698643e-05,
"loss": 1.4251,
"step": 1020
},
{
"epoch": 0.05,
"grad_norm": 0.38146352767944336,
"learning_rate": 4.992465392614345e-05,
"loss": 1.4106,
"step": 1030
},
{
"epoch": 0.05,
"grad_norm": 0.41285020112991333,
"learning_rate": 4.992318454681102e-05,
"loss": 1.3847,
"step": 1040
},
{
"epoch": 0.05,
"grad_norm": 0.38584575057029724,
"learning_rate": 4.992170099982444e-05,
"loss": 1.4685,
"step": 1050
},
{
"epoch": 0.05,
"grad_norm": 0.38032853603363037,
"learning_rate": 4.9920203286027e-05,
"loss": 1.4234,
"step": 1060
},
{
"epoch": 0.05,
"grad_norm": 0.41006794571876526,
"learning_rate": 4.991869140627011e-05,
"loss": 1.4502,
"step": 1070
},
{
"epoch": 0.05,
"grad_norm": 0.35510504245758057,
"learning_rate": 4.99171653614132e-05,
"loss": 1.4633,
"step": 1080
},
{
"epoch": 0.05,
"grad_norm": 0.43953752517700195,
"learning_rate": 4.991562515232374e-05,
"loss": 1.4428,
"step": 1090
},
{
"epoch": 0.05,
"grad_norm": 0.5361181497573853,
"learning_rate": 4.991407077987728e-05,
"loss": 1.3718,
"step": 1100
},
{
"epoch": 0.05,
"grad_norm": 0.43254363536834717,
"learning_rate": 4.9912502244957394e-05,
"loss": 1.458,
"step": 1110
},
{
"epoch": 0.05,
"grad_norm": 0.41394278407096863,
"learning_rate": 4.991091954845573e-05,
"loss": 1.3596,
"step": 1120
},
{
"epoch": 0.05,
"grad_norm": 0.4004674255847931,
"learning_rate": 4.990932269127198e-05,
"loss": 1.4649,
"step": 1130
},
{
"epoch": 0.05,
"grad_norm": 0.4032740592956543,
"learning_rate": 4.990771167431386e-05,
"loss": 1.4114,
"step": 1140
},
{
"epoch": 0.06,
"grad_norm": 0.39824047684669495,
"learning_rate": 4.990608649849719e-05,
"loss": 1.4036,
"step": 1150
},
{
"epoch": 0.06,
"grad_norm": 0.4422670006752014,
"learning_rate": 4.990444716474577e-05,
"loss": 1.3804,
"step": 1160
},
{
"epoch": 0.06,
"grad_norm": 0.44765791296958923,
"learning_rate": 4.990279367399151e-05,
"loss": 1.4613,
"step": 1170
},
{
"epoch": 0.06,
"grad_norm": 0.4563741087913513,
"learning_rate": 4.9901126027174335e-05,
"loss": 1.391,
"step": 1180
},
{
"epoch": 0.06,
"grad_norm": 0.4136991798877716,
"learning_rate": 4.989944422524222e-05,
"loss": 1.4918,
"step": 1190
},
{
"epoch": 0.06,
"grad_norm": 0.46205201745033264,
"learning_rate": 4.989774826915119e-05,
"loss": 1.4146,
"step": 1200
},
{
"epoch": 0.06,
"grad_norm": 0.411703497171402,
"learning_rate": 4.989603815986532e-05,
"loss": 1.3812,
"step": 1210
},
{
"epoch": 0.06,
"grad_norm": 0.43568456172943115,
"learning_rate": 4.989431389835672e-05,
"loss": 1.4803,
"step": 1220
},
{
"epoch": 0.06,
"grad_norm": 0.4594247043132782,
"learning_rate": 4.989257548560557e-05,
"loss": 1.5352,
"step": 1230
},
{
"epoch": 0.06,
"grad_norm": 0.43214985728263855,
"learning_rate": 4.989082292260005e-05,
"loss": 1.4698,
"step": 1240
},
{
"epoch": 0.06,
"grad_norm": 0.3954150676727295,
"learning_rate": 4.9889056210336424e-05,
"loss": 1.4778,
"step": 1250
},
{
"epoch": 0.06,
"grad_norm": 0.457227498292923,
"learning_rate": 4.988727534981898e-05,
"loss": 1.3989,
"step": 1260
},
{
"epoch": 0.06,
"grad_norm": 0.37045395374298096,
"learning_rate": 4.988548034206005e-05,
"loss": 1.4509,
"step": 1270
},
{
"epoch": 0.06,
"grad_norm": 0.5816076993942261,
"learning_rate": 4.988367118808002e-05,
"loss": 1.4639,
"step": 1280
},
{
"epoch": 0.06,
"grad_norm": 0.3573148548603058,
"learning_rate": 4.9881847888907304e-05,
"loss": 1.4041,
"step": 1290
},
{
"epoch": 0.06,
"grad_norm": 0.4649238586425781,
"learning_rate": 4.988001044557836e-05,
"loss": 1.4029,
"step": 1300
},
{
"epoch": 0.06,
"grad_norm": 0.40095779299736023,
"learning_rate": 4.987815885913769e-05,
"loss": 1.3866,
"step": 1310
},
{
"epoch": 0.06,
"grad_norm": 0.4353705048561096,
"learning_rate": 4.987629313063784e-05,
"loss": 1.3828,
"step": 1320
},
{
"epoch": 0.06,
"grad_norm": 0.3809072971343994,
"learning_rate": 4.987441326113938e-05,
"loss": 1.4196,
"step": 1330
},
{
"epoch": 0.06,
"grad_norm": 0.44233348965644836,
"learning_rate": 4.987251925171092e-05,
"loss": 1.4044,
"step": 1340
},
{
"epoch": 0.06,
"grad_norm": 0.6259246468544006,
"learning_rate": 4.987061110342912e-05,
"loss": 1.4262,
"step": 1350
},
{
"epoch": 0.07,
"grad_norm": 0.4725249111652374,
"learning_rate": 4.986868881737869e-05,
"loss": 1.3883,
"step": 1360
},
{
"epoch": 0.07,
"grad_norm": 0.3717879354953766,
"learning_rate": 4.986675239465234e-05,
"loss": 1.4873,
"step": 1370
},
{
"epoch": 0.07,
"grad_norm": 0.47752645611763,
"learning_rate": 4.986480183635084e-05,
"loss": 1.4499,
"step": 1380
},
{
"epoch": 0.07,
"grad_norm": 0.630486011505127,
"learning_rate": 4.986283714358299e-05,
"loss": 1.4473,
"step": 1390
},
{
"epoch": 0.07,
"grad_norm": 0.6709036231040955,
"learning_rate": 4.986085831746562e-05,
"loss": 1.3907,
"step": 1400
},
{
"epoch": 0.07,
"grad_norm": 0.410537451505661,
"learning_rate": 4.985886535912361e-05,
"loss": 1.3758,
"step": 1410
},
{
"epoch": 0.07,
"grad_norm": 0.43040692806243896,
"learning_rate": 4.985685826968985e-05,
"loss": 1.3647,
"step": 1420
},
{
"epoch": 0.07,
"grad_norm": 0.49131736159324646,
"learning_rate": 4.985483705030529e-05,
"loss": 1.4293,
"step": 1430
},
{
"epoch": 0.07,
"grad_norm": 0.43018704652786255,
"learning_rate": 4.985280170211889e-05,
"loss": 1.49,
"step": 1440
},
{
"epoch": 0.07,
"grad_norm": 0.39664191007614136,
"learning_rate": 4.9850752226287646e-05,
"loss": 1.373,
"step": 1450
},
{
"epoch": 0.07,
"grad_norm": 0.4041331708431244,
"learning_rate": 4.984868862397659e-05,
"loss": 1.4126,
"step": 1460
},
{
"epoch": 0.07,
"grad_norm": 0.378964900970459,
"learning_rate": 4.984661089635879e-05,
"loss": 1.3968,
"step": 1470
},
{
"epoch": 0.07,
"grad_norm": 0.4356936812400818,
"learning_rate": 4.984451904461533e-05,
"loss": 1.4776,
"step": 1480
},
{
"epoch": 0.07,
"grad_norm": 0.40792277455329895,
"learning_rate": 4.9842413069935334e-05,
"loss": 1.417,
"step": 1490
},
{
"epoch": 0.07,
"grad_norm": 0.4152584671974182,
"learning_rate": 4.984029297351595e-05,
"loss": 1.4391,
"step": 1500
},
{
"epoch": 0.07,
"grad_norm": 0.4692898988723755,
"learning_rate": 4.983815875656234e-05,
"loss": 1.5064,
"step": 1510
},
{
"epoch": 0.07,
"grad_norm": 0.42290377616882324,
"learning_rate": 4.98360104202877e-05,
"loss": 1.4772,
"step": 1520
},
{
"epoch": 0.07,
"grad_norm": 0.3567737638950348,
"learning_rate": 4.9833847965913295e-05,
"loss": 1.4011,
"step": 1530
},
{
"epoch": 0.07,
"grad_norm": 0.4090891480445862,
"learning_rate": 4.983167139466835e-05,
"loss": 1.3505,
"step": 1540
},
{
"epoch": 0.07,
"grad_norm": 0.48744046688079834,
"learning_rate": 4.9829480707790136e-05,
"loss": 1.4405,
"step": 1550
},
{
"epoch": 0.07,
"grad_norm": 0.33889856934547424,
"learning_rate": 4.982727590652397e-05,
"loss": 1.4648,
"step": 1560
},
{
"epoch": 0.08,
"grad_norm": 0.4974396824836731,
"learning_rate": 4.982505699212318e-05,
"loss": 1.4585,
"step": 1570
},
{
"epoch": 0.08,
"grad_norm": 0.5973418951034546,
"learning_rate": 4.9822823965849094e-05,
"loss": 1.4092,
"step": 1580
},
{
"epoch": 0.08,
"grad_norm": 0.43996697664260864,
"learning_rate": 4.982057682897111e-05,
"loss": 1.4175,
"step": 1590
},
{
"epoch": 0.08,
"grad_norm": 0.3844873905181885,
"learning_rate": 4.981831558276661e-05,
"loss": 1.4742,
"step": 1600
},
{
"epoch": 0.08,
"grad_norm": 0.4250548779964447,
"learning_rate": 4.9816040228520986e-05,
"loss": 1.3715,
"step": 1610
},
{
"epoch": 0.08,
"grad_norm": 0.460089772939682,
"learning_rate": 4.98137507675277e-05,
"loss": 1.3797,
"step": 1620
},
{
"epoch": 0.08,
"grad_norm": 0.4590339958667755,
"learning_rate": 4.981144720108818e-05,
"loss": 1.4128,
"step": 1630
},
{
"epoch": 0.08,
"grad_norm": 0.4135112464427948,
"learning_rate": 4.98091295305119e-05,
"loss": 1.4287,
"step": 1640
},
{
"epoch": 0.08,
"grad_norm": 0.49125537276268005,
"learning_rate": 4.980679775711635e-05,
"loss": 1.4535,
"step": 1650
},
{
"epoch": 0.08,
"grad_norm": 0.36171460151672363,
"learning_rate": 4.980445188222703e-05,
"loss": 1.4254,
"step": 1660
},
{
"epoch": 0.08,
"grad_norm": 0.4327664375305176,
"learning_rate": 4.980209190717747e-05,
"loss": 1.4369,
"step": 1670
},
{
"epoch": 0.08,
"grad_norm": 0.4476342797279358,
"learning_rate": 4.979971783330919e-05,
"loss": 1.4164,
"step": 1680
},
{
"epoch": 0.08,
"grad_norm": 0.4774191081523895,
"learning_rate": 4.9797329661971737e-05,
"loss": 1.4177,
"step": 1690
},
{
"epoch": 0.08,
"grad_norm": 0.4090128540992737,
"learning_rate": 4.9794927394522685e-05,
"loss": 1.3658,
"step": 1700
},
{
"epoch": 0.08,
"grad_norm": 0.4297221899032593,
"learning_rate": 4.979251103232761e-05,
"loss": 1.386,
"step": 1710
},
{
"epoch": 0.08,
"grad_norm": 0.4570232033729553,
"learning_rate": 4.979008057676009e-05,
"loss": 1.3867,
"step": 1720
},
{
"epoch": 0.08,
"grad_norm": 0.4124417304992676,
"learning_rate": 4.9787636029201735e-05,
"loss": 1.4353,
"step": 1730
},
{
"epoch": 0.08,
"grad_norm": 0.42019960284233093,
"learning_rate": 4.978517739104215e-05,
"loss": 1.4863,
"step": 1740
},
{
"epoch": 0.08,
"grad_norm": 0.5385653376579285,
"learning_rate": 4.978270466367895e-05,
"loss": 1.4826,
"step": 1750
},
{
"epoch": 0.08,
"grad_norm": 0.4962645471096039,
"learning_rate": 4.9780217848517764e-05,
"loss": 1.4566,
"step": 1760
},
{
"epoch": 0.08,
"grad_norm": 0.4264776408672333,
"learning_rate": 4.9777716946972244e-05,
"loss": 1.4124,
"step": 1770
},
{
"epoch": 0.09,
"grad_norm": 0.35850468277931213,
"learning_rate": 4.9775201960464015e-05,
"loss": 1.3535,
"step": 1780
},
{
"epoch": 0.09,
"grad_norm": 0.4027429223060608,
"learning_rate": 4.977267289042274e-05,
"loss": 1.4196,
"step": 1790
},
{
"epoch": 0.09,
"grad_norm": 0.40153437852859497,
"learning_rate": 4.977012973828607e-05,
"loss": 1.3722,
"step": 1800
},
{
"epoch": 0.09,
"grad_norm": 0.5421766042709351,
"learning_rate": 4.976757250549968e-05,
"loss": 1.3688,
"step": 1810
},
{
"epoch": 0.09,
"grad_norm": 0.43894025683403015,
"learning_rate": 4.976500119351721e-05,
"loss": 1.4338,
"step": 1820
},
{
"epoch": 0.09,
"grad_norm": 0.46076783537864685,
"learning_rate": 4.9762415803800356e-05,
"loss": 1.4063,
"step": 1830
},
{
"epoch": 0.09,
"grad_norm": 0.3808012306690216,
"learning_rate": 4.975981633781878e-05,
"loss": 1.436,
"step": 1840
},
{
"epoch": 0.09,
"grad_norm": 0.4237988293170929,
"learning_rate": 4.9757202797050154e-05,
"loss": 1.4772,
"step": 1850
},
{
"epoch": 0.09,
"grad_norm": 0.45093056559562683,
"learning_rate": 4.975457518298015e-05,
"loss": 1.4557,
"step": 1860
},
{
"epoch": 0.09,
"grad_norm": 0.5528602600097656,
"learning_rate": 4.975193349710245e-05,
"loss": 1.3448,
"step": 1870
},
{
"epoch": 0.09,
"grad_norm": 0.32752522826194763,
"learning_rate": 4.974927774091872e-05,
"loss": 1.3241,
"step": 1880
},
{
"epoch": 0.09,
"grad_norm": 0.48677608370780945,
"learning_rate": 4.9746607915938645e-05,
"loss": 1.4538,
"step": 1890
},
{
"epoch": 0.09,
"grad_norm": 0.5248567461967468,
"learning_rate": 4.974392402367988e-05,
"loss": 1.3626,
"step": 1900
},
{
"epoch": 0.09,
"grad_norm": 0.3645094633102417,
"learning_rate": 4.9741226065668095e-05,
"loss": 1.3958,
"step": 1910
},
{
"epoch": 0.09,
"grad_norm": 0.3824084401130676,
"learning_rate": 4.973851404343696e-05,
"loss": 1.5136,
"step": 1920
},
{
"epoch": 0.09,
"grad_norm": 0.40341123938560486,
"learning_rate": 4.9735787958528115e-05,
"loss": 1.432,
"step": 1930
},
{
"epoch": 0.09,
"grad_norm": 0.44214069843292236,
"learning_rate": 4.973304781249123e-05,
"loss": 1.3792,
"step": 1940
},
{
"epoch": 0.09,
"grad_norm": 0.49846524000167847,
"learning_rate": 4.973029360688393e-05,
"loss": 1.3372,
"step": 1950
},
{
"epoch": 0.09,
"grad_norm": 0.3863857090473175,
"learning_rate": 4.972752534327187e-05,
"loss": 1.4074,
"step": 1960
},
{
"epoch": 0.09,
"grad_norm": 0.3839406967163086,
"learning_rate": 4.972474302322866e-05,
"loss": 1.3545,
"step": 1970
},
{
"epoch": 0.1,
"grad_norm": 0.40729382634162903,
"learning_rate": 4.972194664833593e-05,
"loss": 1.3504,
"step": 1980
},
{
"epoch": 0.1,
"grad_norm": 0.4500448703765869,
"learning_rate": 4.971913622018328e-05,
"loss": 1.462,
"step": 1990
},
{
"epoch": 0.1,
"grad_norm": 0.4089118540287018,
"learning_rate": 4.971631174036832e-05,
"loss": 1.4078,
"step": 2000
},
{
"epoch": 0.1,
"grad_norm": 0.568731963634491,
"learning_rate": 4.97134732104966e-05,
"loss": 1.4429,
"step": 2010
},
{
"epoch": 0.1,
"grad_norm": 0.4214164912700653,
"learning_rate": 4.9710620632181725e-05,
"loss": 1.382,
"step": 2020
},
{
"epoch": 0.1,
"grad_norm": 0.44208085536956787,
"learning_rate": 4.970775400704524e-05,
"loss": 1.4608,
"step": 2030
},
{
"epoch": 0.1,
"grad_norm": 0.4478038251399994,
"learning_rate": 4.9704873336716676e-05,
"loss": 1.4018,
"step": 2040
},
{
"epoch": 0.1,
"grad_norm": 0.41761326789855957,
"learning_rate": 4.970197862283357e-05,
"loss": 1.3665,
"step": 2050
},
{
"epoch": 0.1,
"grad_norm": 0.3954784572124481,
"learning_rate": 4.969906986704143e-05,
"loss": 1.373,
"step": 2060
},
{
"epoch": 0.1,
"grad_norm": 0.42724207043647766,
"learning_rate": 4.9696147070993736e-05,
"loss": 1.4079,
"step": 2070
},
{
"epoch": 0.1,
"grad_norm": 0.5061061382293701,
"learning_rate": 4.9693210236351975e-05,
"loss": 1.443,
"step": 2080
},
{
"epoch": 0.1,
"grad_norm": 0.3955136835575104,
"learning_rate": 4.969025936478558e-05,
"loss": 1.3522,
"step": 2090
},
{
"epoch": 0.1,
"grad_norm": 0.3537675738334656,
"learning_rate": 4.9687294457972e-05,
"loss": 1.3961,
"step": 2100
},
{
"epoch": 0.1,
"grad_norm": 0.3633611798286438,
"learning_rate": 4.968431551759664e-05,
"loss": 1.3743,
"step": 2110
},
{
"epoch": 0.1,
"grad_norm": 0.4026493430137634,
"learning_rate": 4.968132254535288e-05,
"loss": 1.4974,
"step": 2120
},
{
"epoch": 0.1,
"grad_norm": 0.4304758608341217,
"learning_rate": 4.967831554294209e-05,
"loss": 1.3866,
"step": 2130
},
{
"epoch": 0.1,
"grad_norm": 0.318921834230423,
"learning_rate": 4.96752945120736e-05,
"loss": 1.5227,
"step": 2140
},
{
"epoch": 0.1,
"grad_norm": 0.38373032212257385,
"learning_rate": 4.967225945446474e-05,
"loss": 1.3768,
"step": 2150
},
{
"epoch": 0.1,
"grad_norm": 0.5846730470657349,
"learning_rate": 4.966921037184078e-05,
"loss": 1.4046,
"step": 2160
},
{
"epoch": 0.1,
"grad_norm": 0.4191892147064209,
"learning_rate": 4.966614726593499e-05,
"loss": 1.3858,
"step": 2170
},
{
"epoch": 0.1,
"grad_norm": 0.5004653930664062,
"learning_rate": 4.96630701384886e-05,
"loss": 1.4205,
"step": 2180
},
{
"epoch": 0.11,
"grad_norm": 0.4502711296081543,
"learning_rate": 4.96599789912508e-05,
"loss": 1.4226,
"step": 2190
},
{
"epoch": 0.11,
"grad_norm": 0.4821077883243561,
"learning_rate": 4.965687382597878e-05,
"loss": 1.4797,
"step": 2200
},
{
"epoch": 0.11,
"grad_norm": 0.41166359186172485,
"learning_rate": 4.9653754644437665e-05,
"loss": 1.3957,
"step": 2210
},
{
"epoch": 0.11,
"grad_norm": 0.3595781922340393,
"learning_rate": 4.9650621448400566e-05,
"loss": 1.4692,
"step": 2220
},
{
"epoch": 0.11,
"grad_norm": 0.361824631690979,
"learning_rate": 4.964747423964857e-05,
"loss": 1.3711,
"step": 2230
},
{
"epoch": 0.11,
"grad_norm": 0.4380735456943512,
"learning_rate": 4.96443130199707e-05,
"loss": 1.3662,
"step": 2240
},
{
"epoch": 0.11,
"grad_norm": 0.4048125743865967,
"learning_rate": 4.964113779116397e-05,
"loss": 1.4251,
"step": 2250
},
{
"epoch": 0.11,
"grad_norm": 0.42301908135414124,
"learning_rate": 4.963794855503334e-05,
"loss": 1.4315,
"step": 2260
},
{
"epoch": 0.11,
"grad_norm": 0.3762255609035492,
"learning_rate": 4.963474531339174e-05,
"loss": 1.461,
"step": 2270
},
{
"epoch": 0.11,
"grad_norm": 0.40281182527542114,
"learning_rate": 4.963152806806007e-05,
"loss": 1.4259,
"step": 2280
},
{
"epoch": 0.11,
"grad_norm": 0.32970255613327026,
"learning_rate": 4.962829682086718e-05,
"loss": 1.4,
"step": 2290
},
{
"epoch": 0.11,
"grad_norm": 0.5229349136352539,
"learning_rate": 4.9625051573649885e-05,
"loss": 1.4102,
"step": 2300
},
{
"epoch": 0.11,
"grad_norm": 0.37456557154655457,
"learning_rate": 4.962179232825294e-05,
"loss": 1.3944,
"step": 2310
},
{
"epoch": 0.11,
"grad_norm": 0.4639645516872406,
"learning_rate": 4.961851908652908e-05,
"loss": 1.4583,
"step": 2320
},
{
"epoch": 0.11,
"grad_norm": 0.46145012974739075,
"learning_rate": 4.9615231850338995e-05,
"loss": 1.3651,
"step": 2330
},
{
"epoch": 0.11,
"grad_norm": 0.40316319465637207,
"learning_rate": 4.9611930621551314e-05,
"loss": 1.3492,
"step": 2340
},
{
"epoch": 0.11,
"grad_norm": 0.44597890973091125,
"learning_rate": 4.960894755352231e-05,
"loss": 1.4328,
"step": 2350
},
{
"epoch": 0.11,
"grad_norm": 0.43351566791534424,
"learning_rate": 4.9605619743975794e-05,
"loss": 1.3636,
"step": 2360
},
{
"epoch": 0.11,
"grad_norm": 0.3685530722141266,
"learning_rate": 4.960227794729572e-05,
"loss": 1.4333,
"step": 2370
},
{
"epoch": 0.11,
"grad_norm": 0.4043569564819336,
"learning_rate": 4.959892216538174e-05,
"loss": 1.3665,
"step": 2380
},
{
"epoch": 0.11,
"grad_norm": 0.42655327916145325,
"learning_rate": 4.9595552400141446e-05,
"loss": 1.4345,
"step": 2390
},
{
"epoch": 0.12,
"grad_norm": 0.4936932623386383,
"learning_rate": 4.95921686534904e-05,
"loss": 1.3769,
"step": 2400
},
{
"epoch": 0.12,
"grad_norm": 0.4691634774208069,
"learning_rate": 4.95887709273521e-05,
"loss": 1.3498,
"step": 2410
},
{
"epoch": 0.12,
"grad_norm": 0.35662075877189636,
"learning_rate": 4.9585359223658e-05,
"loss": 1.4393,
"step": 2420
},
{
"epoch": 0.12,
"grad_norm": 0.3932383358478546,
"learning_rate": 4.958193354434749e-05,
"loss": 1.4553,
"step": 2430
},
{
"epoch": 0.12,
"grad_norm": 0.38411515951156616,
"learning_rate": 4.95784938913679e-05,
"loss": 1.3984,
"step": 2440
},
{
"epoch": 0.12,
"grad_norm": 0.4691387712955475,
"learning_rate": 4.957504026667451e-05,
"loss": 1.3944,
"step": 2450
},
{
"epoch": 0.12,
"grad_norm": 0.47951576113700867,
"learning_rate": 4.9571572672230556e-05,
"loss": 1.3374,
"step": 2460
},
{
"epoch": 0.12,
"grad_norm": 0.35057947039604187,
"learning_rate": 4.9568091110007184e-05,
"loss": 1.5118,
"step": 2470
},
{
"epoch": 0.12,
"grad_norm": 0.4169234335422516,
"learning_rate": 4.956459558198351e-05,
"loss": 1.3027,
"step": 2480
},
{
"epoch": 0.12,
"grad_norm": 0.4908248782157898,
"learning_rate": 4.956108609014658e-05,
"loss": 1.4735,
"step": 2490
},
{
"epoch": 0.12,
"grad_norm": 0.3674967586994171,
"learning_rate": 4.955756263649136e-05,
"loss": 1.4357,
"step": 2500
},
{
"epoch": 0.12,
"grad_norm": 0.479251503944397,
"learning_rate": 4.955402522302078e-05,
"loss": 1.431,
"step": 2510
},
{
"epoch": 0.12,
"grad_norm": 0.4014340937137604,
"learning_rate": 4.955047385174568e-05,
"loss": 1.2962,
"step": 2520
},
{
"epoch": 0.12,
"grad_norm": 0.4915308952331543,
"learning_rate": 4.954690852468486e-05,
"loss": 1.4719,
"step": 2530
},
{
"epoch": 0.12,
"grad_norm": 0.4371248185634613,
"learning_rate": 4.9543329243865044e-05,
"loss": 1.3533,
"step": 2540
},
{
"epoch": 0.12,
"grad_norm": 0.4188721179962158,
"learning_rate": 4.953973601132086e-05,
"loss": 1.4621,
"step": 2550
},
{
"epoch": 0.12,
"grad_norm": 0.5254568457603455,
"learning_rate": 4.9536128829094915e-05,
"loss": 1.431,
"step": 2560
},
{
"epoch": 0.12,
"grad_norm": 0.4436979293823242,
"learning_rate": 4.9532507699237705e-05,
"loss": 1.449,
"step": 2570
},
{
"epoch": 0.12,
"grad_norm": 0.4138881266117096,
"learning_rate": 4.952887262380769e-05,
"loss": 1.4137,
"step": 2580
},
{
"epoch": 0.12,
"grad_norm": 0.3862578272819519,
"learning_rate": 4.9525223604871216e-05,
"loss": 1.4176,
"step": 2590
},
{
"epoch": 0.12,
"grad_norm": 0.4499189555644989,
"learning_rate": 4.95215606445026e-05,
"loss": 1.388,
"step": 2600
},
{
"epoch": 0.13,
"grad_norm": 0.34604617953300476,
"learning_rate": 4.951788374478405e-05,
"loss": 1.3782,
"step": 2610
},
{
"epoch": 0.13,
"grad_norm": 0.46878838539123535,
"learning_rate": 4.951419290780571e-05,
"loss": 1.432,
"step": 2620
},
{
"epoch": 0.13,
"grad_norm": 0.44375401735305786,
"learning_rate": 4.9510488135665656e-05,
"loss": 1.4306,
"step": 2630
},
{
"epoch": 0.13,
"grad_norm": 0.4289904832839966,
"learning_rate": 4.950676943046986e-05,
"loss": 1.3706,
"step": 2640
},
{
"epoch": 0.13,
"grad_norm": 0.3851960003376007,
"learning_rate": 4.950303679433225e-05,
"loss": 1.38,
"step": 2650
},
{
"epoch": 0.13,
"grad_norm": 0.3767862021923065,
"learning_rate": 4.9499290229374626e-05,
"loss": 1.415,
"step": 2660
},
{
"epoch": 0.13,
"grad_norm": 0.44467443227767944,
"learning_rate": 4.949552973772677e-05,
"loss": 1.3988,
"step": 2670
},
{
"epoch": 0.13,
"grad_norm": 0.3889351785182953,
"learning_rate": 4.949175532152631e-05,
"loss": 1.5051,
"step": 2680
},
{
"epoch": 0.13,
"grad_norm": 0.3801763355731964,
"learning_rate": 4.948796698291884e-05,
"loss": 1.4756,
"step": 2690
},
{
"epoch": 0.13,
"grad_norm": 0.34679311513900757,
"learning_rate": 4.9484164724057856e-05,
"loss": 1.4772,
"step": 2700
},
{
"epoch": 0.13,
"grad_norm": 0.5637999176979065,
"learning_rate": 4.9480348547104746e-05,
"loss": 1.4575,
"step": 2710
},
{
"epoch": 0.13,
"grad_norm": 0.4573119282722473,
"learning_rate": 4.947651845422885e-05,
"loss": 1.4303,
"step": 2720
},
{
"epoch": 0.13,
"grad_norm": 0.4247356951236725,
"learning_rate": 4.947267444760737e-05,
"loss": 1.359,
"step": 2730
},
{
"epoch": 0.13,
"grad_norm": 0.5523324012756348,
"learning_rate": 4.946881652942547e-05,
"loss": 1.4418,
"step": 2740
},
{
"epoch": 0.13,
"grad_norm": 0.40322351455688477,
"learning_rate": 4.946494470187616e-05,
"loss": 1.4137,
"step": 2750
},
{
"epoch": 0.13,
"grad_norm": 0.44482749700546265,
"learning_rate": 4.946105896716042e-05,
"loss": 1.4511,
"step": 2760
},
{
"epoch": 0.13,
"grad_norm": 0.3632228970527649,
"learning_rate": 4.945715932748709e-05,
"loss": 1.3763,
"step": 2770
},
{
"epoch": 0.13,
"grad_norm": 0.4271148443222046,
"learning_rate": 4.945324578507294e-05,
"loss": 1.3768,
"step": 2780
},
{
"epoch": 0.13,
"grad_norm": 0.43617910146713257,
"learning_rate": 4.944931834214264e-05,
"loss": 1.5105,
"step": 2790
},
{
"epoch": 0.13,
"grad_norm": 0.4843004643917084,
"learning_rate": 4.944537700092874e-05,
"loss": 1.4622,
"step": 2800
},
{
"epoch": 0.13,
"grad_norm": 0.4292265474796295,
"learning_rate": 4.944142176367171e-05,
"loss": 1.4363,
"step": 2810
},
{
"epoch": 0.14,
"grad_norm": 0.3804440200328827,
"learning_rate": 4.943745263261993e-05,
"loss": 1.424,
"step": 2820
},
{
"epoch": 0.14,
"grad_norm": 0.4906880557537079,
"learning_rate": 4.9433868537343445e-05,
"loss": 1.4086,
"step": 2830
},
{
"epoch": 0.14,
"grad_norm": 0.3771170675754547,
"learning_rate": 4.942987301430415e-05,
"loss": 1.4359,
"step": 2840
},
{
"epoch": 0.14,
"grad_norm": 0.46057745814323425,
"learning_rate": 4.942586360403501e-05,
"loss": 1.4257,
"step": 2850
},
{
"epoch": 0.14,
"grad_norm": 0.3592281639575958,
"learning_rate": 4.942184030881519e-05,
"loss": 1.4595,
"step": 2860
},
{
"epoch": 0.14,
"grad_norm": 0.40268605947494507,
"learning_rate": 4.941780313093175e-05,
"loss": 1.3317,
"step": 2870
},
{
"epoch": 0.14,
"grad_norm": 0.3508237898349762,
"learning_rate": 4.941375207267962e-05,
"loss": 1.4095,
"step": 2880
},
{
"epoch": 0.14,
"grad_norm": 0.39787015318870544,
"learning_rate": 4.9409687136361645e-05,
"loss": 1.3912,
"step": 2890
},
{
"epoch": 0.14,
"grad_norm": 0.4719141721725464,
"learning_rate": 4.940560832428856e-05,
"loss": 1.411,
"step": 2900
},
{
"epoch": 0.14,
"grad_norm": 0.4503662586212158,
"learning_rate": 4.940151563877895e-05,
"loss": 1.3731,
"step": 2910
},
{
"epoch": 0.14,
"grad_norm": 0.42547717690467834,
"learning_rate": 4.939740908215934e-05,
"loss": 1.3869,
"step": 2920
},
{
"epoch": 0.14,
"grad_norm": 0.3675297796726227,
"learning_rate": 4.939328865676409e-05,
"loss": 1.3312,
"step": 2930
},
{
"epoch": 0.14,
"grad_norm": 0.3596685230731964,
"learning_rate": 4.9389154364935484e-05,
"loss": 1.3562,
"step": 2940
},
{
"epoch": 0.14,
"grad_norm": 0.4983693063259125,
"learning_rate": 4.9385006209023674e-05,
"loss": 1.4251,
"step": 2950
},
{
"epoch": 0.14,
"grad_norm": 0.3797384202480316,
"learning_rate": 4.9380844191386675e-05,
"loss": 1.3221,
"step": 2960
},
{
"epoch": 0.14,
"grad_norm": 0.376395583152771,
"learning_rate": 4.937666831439042e-05,
"loss": 1.358,
"step": 2970
},
{
"epoch": 0.14,
"grad_norm": 0.4990999400615692,
"learning_rate": 4.937247858040867e-05,
"loss": 1.4953,
"step": 2980
},
{
"epoch": 0.14,
"grad_norm": 0.35752540826797485,
"learning_rate": 4.9368274991823114e-05,
"loss": 1.4265,
"step": 2990
},
{
"epoch": 0.14,
"grad_norm": 0.3814866244792938,
"learning_rate": 4.936405755102328e-05,
"loss": 1.4071,
"step": 3000
},
{
"epoch": 0.14,
"grad_norm": 0.3756539225578308,
"learning_rate": 4.93598262604066e-05,
"loss": 1.4517,
"step": 3010
},
{
"epoch": 0.14,
"grad_norm": 0.30514612793922424,
"learning_rate": 4.9355581122378345e-05,
"loss": 1.4617,
"step": 3020
},
{
"epoch": 0.15,
"grad_norm": 0.33860358595848083,
"learning_rate": 4.935132213935169e-05,
"loss": 1.3517,
"step": 3030
},
{
"epoch": 0.15,
"grad_norm": 0.3979358375072479,
"learning_rate": 4.9347049313747645e-05,
"loss": 1.3711,
"step": 3040
},
{
"epoch": 0.15,
"grad_norm": 0.4487731158733368,
"learning_rate": 4.9342762647995133e-05,
"loss": 1.3902,
"step": 3050
},
{
"epoch": 0.15,
"grad_norm": 0.3981623649597168,
"learning_rate": 4.933846214453091e-05,
"loss": 1.4817,
"step": 3060
},
{
"epoch": 0.15,
"grad_norm": 0.433854877948761,
"learning_rate": 4.9334147805799615e-05,
"loss": 1.4142,
"step": 3070
},
{
"epoch": 0.15,
"grad_norm": 0.42882010340690613,
"learning_rate": 4.932981963425374e-05,
"loss": 1.3652,
"step": 3080
},
{
"epoch": 0.15,
"grad_norm": 0.391488641500473,
"learning_rate": 4.9325477632353644e-05,
"loss": 1.4538,
"step": 3090
},
{
"epoch": 0.15,
"grad_norm": 0.38359108567237854,
"learning_rate": 4.932112180256756e-05,
"loss": 1.3608,
"step": 3100
},
{
"epoch": 0.15,
"grad_norm": 0.482505738735199,
"learning_rate": 4.9316752147371567e-05,
"loss": 1.3986,
"step": 3110
},
{
"epoch": 0.15,
"grad_norm": 0.45531147718429565,
"learning_rate": 4.931236866924961e-05,
"loss": 1.4223,
"step": 3120
},
{
"epoch": 0.15,
"grad_norm": 0.37000513076782227,
"learning_rate": 4.9307971370693477e-05,
"loss": 1.4038,
"step": 3130
},
{
"epoch": 0.15,
"grad_norm": 0.3654976487159729,
"learning_rate": 4.930356025420284e-05,
"loss": 1.3829,
"step": 3140
},
{
"epoch": 0.15,
"grad_norm": 0.43360087275505066,
"learning_rate": 4.929913532228521e-05,
"loss": 1.3588,
"step": 3150
},
{
"epoch": 0.15,
"grad_norm": 0.3908851444721222,
"learning_rate": 4.929469657745593e-05,
"loss": 1.368,
"step": 3160
},
{
"epoch": 0.15,
"grad_norm": 0.3879352807998657,
"learning_rate": 4.929024402223825e-05,
"loss": 1.3583,
"step": 3170
},
{
"epoch": 0.15,
"grad_norm": 0.4229816198348999,
"learning_rate": 4.9285777659163213e-05,
"loss": 1.3706,
"step": 3180
},
{
"epoch": 0.15,
"grad_norm": 0.3288842439651489,
"learning_rate": 4.928129749076975e-05,
"loss": 1.414,
"step": 3190
},
{
"epoch": 0.15,
"grad_norm": 0.4697210490703583,
"learning_rate": 4.9276803519604607e-05,
"loss": 1.4356,
"step": 3200
},
{
"epoch": 0.15,
"grad_norm": 0.45226651430130005,
"learning_rate": 4.927229574822242e-05,
"loss": 1.4269,
"step": 3210
},
{
"epoch": 0.15,
"grad_norm": 0.33699142932891846,
"learning_rate": 4.9267774179185624e-05,
"loss": 1.3898,
"step": 3220
},
{
"epoch": 0.16,
"grad_norm": 0.36085259914398193,
"learning_rate": 4.9263238815064525e-05,
"loss": 1.4703,
"step": 3230
},
{
"epoch": 0.16,
"grad_norm": 0.38384854793548584,
"learning_rate": 4.9258689658437267e-05,
"loss": 1.4042,
"step": 3240
},
{
"epoch": 0.16,
"grad_norm": 0.3826994001865387,
"learning_rate": 4.925412671188983e-05,
"loss": 1.4488,
"step": 3250
},
{
"epoch": 0.16,
"grad_norm": 0.3617662489414215,
"learning_rate": 4.924954997801602e-05,
"loss": 1.3895,
"step": 3260
},
{
"epoch": 0.16,
"grad_norm": 0.39644262194633484,
"learning_rate": 4.9244959459417506e-05,
"loss": 1.4994,
"step": 3270
},
{
"epoch": 0.16,
"grad_norm": 0.4409465789794922,
"learning_rate": 4.924035515870379e-05,
"loss": 1.4745,
"step": 3280
},
{
"epoch": 0.16,
"grad_norm": 0.41163429617881775,
"learning_rate": 4.923573707849218e-05,
"loss": 1.3606,
"step": 3290
},
{
"epoch": 0.16,
"grad_norm": 0.3120229244232178,
"learning_rate": 4.923110522140785e-05,
"loss": 1.3314,
"step": 3300
},
{
"epoch": 0.16,
"grad_norm": 0.37128886580467224,
"learning_rate": 4.922645959008378e-05,
"loss": 1.3961,
"step": 3310
},
{
"epoch": 0.16,
"grad_norm": 0.4180125892162323,
"learning_rate": 4.922180018716081e-05,
"loss": 1.4276,
"step": 3320
},
{
"epoch": 0.16,
"grad_norm": 0.4499339461326599,
"learning_rate": 4.9217127015287577e-05,
"loss": 1.4827,
"step": 3330
},
{
"epoch": 0.16,
"grad_norm": 0.354478657245636,
"learning_rate": 4.921244007712057e-05,
"loss": 1.4417,
"step": 3340
},
{
"epoch": 0.16,
"grad_norm": 0.47401493787765503,
"learning_rate": 4.920773937532408e-05,
"loss": 1.4135,
"step": 3350
},
{
"epoch": 0.16,
"grad_norm": 0.38429275155067444,
"learning_rate": 4.920302491257024e-05,
"loss": 1.4785,
"step": 3360
},
{
"epoch": 0.16,
"grad_norm": 0.413880854845047,
"learning_rate": 4.919829669153899e-05,
"loss": 1.4045,
"step": 3370
},
{
"epoch": 0.16,
"grad_norm": 0.33524903655052185,
"learning_rate": 4.919355471491812e-05,
"loss": 1.3569,
"step": 3380
},
{
"epoch": 0.16,
"grad_norm": 0.3383985459804535,
"learning_rate": 4.9188798985403206e-05,
"loss": 1.4347,
"step": 3390
},
{
"epoch": 0.16,
"grad_norm": 0.34362682700157166,
"learning_rate": 4.918402950569766e-05,
"loss": 1.4035,
"step": 3400
},
{
"epoch": 0.16,
"grad_norm": 0.6409430503845215,
"learning_rate": 4.9179246278512704e-05,
"loss": 1.3909,
"step": 3410
},
{
"epoch": 0.16,
"grad_norm": 0.4362596869468689,
"learning_rate": 4.917444930656737e-05,
"loss": 1.3633,
"step": 3420
},
{
"epoch": 0.16,
"grad_norm": 0.4037545323371887,
"learning_rate": 4.916963859258853e-05,
"loss": 1.3498,
"step": 3430
},
{
"epoch": 0.17,
"grad_norm": 0.44377681612968445,
"learning_rate": 4.916481413931083e-05,
"loss": 1.3695,
"step": 3440
},
{
"epoch": 0.17,
"grad_norm": 0.3952435851097107,
"learning_rate": 4.915997594947675e-05,
"loss": 1.3116,
"step": 3450
},
{
"epoch": 0.17,
"grad_norm": 0.3884615898132324,
"learning_rate": 4.915512402583658e-05,
"loss": 1.4249,
"step": 3460
},
{
"epoch": 0.17,
"grad_norm": 0.4115287661552429,
"learning_rate": 4.915025837114839e-05,
"loss": 1.4529,
"step": 3470
},
{
"epoch": 0.17,
"grad_norm": 0.526142954826355,
"learning_rate": 4.91453789881781e-05,
"loss": 1.3492,
"step": 3480
},
{
"epoch": 0.17,
"grad_norm": 0.451386034488678,
"learning_rate": 4.914048587969939e-05,
"loss": 1.4023,
"step": 3490
},
{
"epoch": 0.17,
"grad_norm": 0.4508513808250427,
"learning_rate": 4.913557904849376e-05,
"loss": 1.3299,
"step": 3500
},
{
"epoch": 0.17,
"grad_norm": 0.3877275288105011,
"learning_rate": 4.913065849735053e-05,
"loss": 1.4062,
"step": 3510
},
{
"epoch": 0.17,
"grad_norm": 0.35273948311805725,
"learning_rate": 4.912572422906679e-05,
"loss": 1.4358,
"step": 3520
},
{
"epoch": 0.17,
"grad_norm": 0.41777125000953674,
"learning_rate": 4.912077624644743e-05,
"loss": 1.4201,
"step": 3530
},
{
"epoch": 0.17,
"grad_norm": 0.4178641140460968,
"learning_rate": 4.911581455230516e-05,
"loss": 1.4423,
"step": 3540
},
{
"epoch": 0.17,
"grad_norm": 0.3660828173160553,
"learning_rate": 4.9110839149460465e-05,
"loss": 1.4171,
"step": 3550
},
{
"epoch": 0.17,
"grad_norm": 0.45600688457489014,
"learning_rate": 4.910585004074162e-05,
"loss": 1.355,
"step": 3560
},
{
"epoch": 0.17,
"grad_norm": 0.3535476326942444,
"learning_rate": 4.91008472289847e-05,
"loss": 1.4513,
"step": 3570
},
{
"epoch": 0.17,
"grad_norm": 0.34211233258247375,
"learning_rate": 4.909583071703358e-05,
"loss": 1.434,
"step": 3580
},
{
"epoch": 0.17,
"grad_norm": 0.4727470576763153,
"learning_rate": 4.909080050773989e-05,
"loss": 1.4231,
"step": 3590
},
{
"epoch": 0.17,
"grad_norm": 0.41787880659103394,
"learning_rate": 4.908575660396308e-05,
"loss": 1.3487,
"step": 3600
},
{
"epoch": 0.17,
"grad_norm": 0.33646029233932495,
"learning_rate": 4.9080699008570364e-05,
"loss": 1.3225,
"step": 3610
},
{
"epoch": 0.17,
"grad_norm": 0.36372923851013184,
"learning_rate": 4.9075627724436755e-05,
"loss": 1.3935,
"step": 3620
},
{
"epoch": 0.17,
"grad_norm": 0.4421246647834778,
"learning_rate": 4.907054275444503e-05,
"loss": 1.448,
"step": 3630
},
{
"epoch": 0.17,
"grad_norm": 0.4159976840019226,
"learning_rate": 4.9065444101485755e-05,
"loss": 1.4128,
"step": 3640
},
{
"epoch": 0.18,
"grad_norm": 0.43933582305908203,
"learning_rate": 4.906033176845729e-05,
"loss": 1.451,
"step": 3650
},
{
"epoch": 0.18,
"grad_norm": 0.36872079968452454,
"learning_rate": 4.9055205758265735e-05,
"loss": 1.3918,
"step": 3660
},
{
"epoch": 0.18,
"grad_norm": 0.34184175729751587,
"learning_rate": 4.9050066073824986e-05,
"loss": 1.4736,
"step": 3670
},
{
"epoch": 0.18,
"grad_norm": 0.3967076241970062,
"learning_rate": 4.904491271805672e-05,
"loss": 1.4204,
"step": 3680
},
{
"epoch": 0.18,
"grad_norm": 0.32891419529914856,
"learning_rate": 4.9039745693890373e-05,
"loss": 1.4435,
"step": 3690
},
{
"epoch": 0.18,
"grad_norm": 0.3899432420730591,
"learning_rate": 4.9034565004263155e-05,
"loss": 1.3658,
"step": 3700
},
{
"epoch": 0.18,
"grad_norm": 0.3919103741645813,
"learning_rate": 4.902937065212004e-05,
"loss": 1.3784,
"step": 3710
},
{
"epoch": 0.18,
"grad_norm": 0.5006517767906189,
"learning_rate": 4.902416264041378e-05,
"loss": 1.4704,
"step": 3720
},
{
"epoch": 0.18,
"grad_norm": 0.33027639985084534,
"learning_rate": 4.901894097210488e-05,
"loss": 1.4366,
"step": 3730
},
{
"epoch": 0.18,
"grad_norm": 0.3469955325126648,
"learning_rate": 4.90137056501616e-05,
"loss": 1.3533,
"step": 3740
},
{
"epoch": 0.18,
"grad_norm": 0.3639344871044159,
"learning_rate": 4.900845667756e-05,
"loss": 1.3578,
"step": 3750
},
{
"epoch": 0.18,
"grad_norm": 0.33176496624946594,
"learning_rate": 4.9003194057283844e-05,
"loss": 1.4319,
"step": 3760
},
{
"epoch": 0.18,
"grad_norm": 0.3801308572292328,
"learning_rate": 4.899791779232471e-05,
"loss": 1.4473,
"step": 3770
},
{
"epoch": 0.18,
"grad_norm": 0.45975393056869507,
"learning_rate": 4.8992627885681885e-05,
"loss": 1.3751,
"step": 3780
},
{
"epoch": 0.18,
"grad_norm": 0.3475228250026703,
"learning_rate": 4.898732434036244e-05,
"loss": 1.3992,
"step": 3790
},
{
"epoch": 0.18,
"grad_norm": 0.4052601158618927,
"learning_rate": 4.898200715938118e-05,
"loss": 1.4318,
"step": 3800
},
{
"epoch": 0.18,
"grad_norm": 0.4561234414577484,
"learning_rate": 4.8976676345760695e-05,
"loss": 1.412,
"step": 3810
},
{
"epoch": 0.18,
"grad_norm": 0.47499555349349976,
"learning_rate": 4.897133190253128e-05,
"loss": 1.435,
"step": 3820
},
{
"epoch": 0.18,
"grad_norm": 0.4745842218399048,
"learning_rate": 4.8965973832731004e-05,
"loss": 1.4243,
"step": 3830
},
{
"epoch": 0.18,
"grad_norm": 0.4241657257080078,
"learning_rate": 4.8960602139405684e-05,
"loss": 1.4967,
"step": 3840
},
{
"epoch": 0.18,
"grad_norm": 0.31881964206695557,
"learning_rate": 4.895521682560886e-05,
"loss": 1.3714,
"step": 3850
},
{
"epoch": 0.19,
"grad_norm": 0.33000463247299194,
"learning_rate": 4.894981789440184e-05,
"loss": 1.3705,
"step": 3860
},
{
"epoch": 0.19,
"grad_norm": 0.4364418685436249,
"learning_rate": 4.894440534885366e-05,
"loss": 1.4108,
"step": 3870
},
{
"epoch": 0.19,
"grad_norm": 0.5331186652183533,
"learning_rate": 4.893897919204109e-05,
"loss": 1.4444,
"step": 3880
},
{
"epoch": 0.19,
"grad_norm": 0.3436141908168793,
"learning_rate": 4.893353942704866e-05,
"loss": 1.3817,
"step": 3890
},
{
"epoch": 0.19,
"grad_norm": 0.361950159072876,
"learning_rate": 4.8928086056968595e-05,
"loss": 1.4403,
"step": 3900
},
{
"epoch": 0.19,
"grad_norm": 0.39649009704589844,
"learning_rate": 4.89226190849009e-05,
"loss": 1.3543,
"step": 3910
},
{
"epoch": 0.19,
"grad_norm": 0.41909700632095337,
"learning_rate": 4.8917138513953284e-05,
"loss": 1.3536,
"step": 3920
},
{
"epoch": 0.19,
"grad_norm": 0.38983240723609924,
"learning_rate": 4.891164434724119e-05,
"loss": 1.3415,
"step": 3930
},
{
"epoch": 0.19,
"grad_norm": 0.44484665989875793,
"learning_rate": 4.89061365878878e-05,
"loss": 1.4124,
"step": 3940
},
{
"epoch": 0.19,
"grad_norm": 0.37146005034446716,
"learning_rate": 4.8900615239024e-05,
"loss": 1.3452,
"step": 3950
},
{
"epoch": 0.19,
"grad_norm": 0.319859117269516,
"learning_rate": 4.889508030378843e-05,
"loss": 1.4567,
"step": 3960
},
{
"epoch": 0.19,
"grad_norm": 0.47834596037864685,
"learning_rate": 4.8889531785327435e-05,
"loss": 1.4006,
"step": 3970
},
{
"epoch": 0.19,
"grad_norm": 0.35665494203567505,
"learning_rate": 4.8883969686795086e-05,
"loss": 1.3657,
"step": 3980
},
{
"epoch": 0.19,
"grad_norm": 0.47481733560562134,
"learning_rate": 4.887839401135318e-05,
"loss": 1.3703,
"step": 3990
},
{
"epoch": 0.19,
"grad_norm": 0.366764098405838,
"learning_rate": 4.887280476217121e-05,
"loss": 1.3724,
"step": 4000
},
{
"epoch": 0.19,
"grad_norm": 0.43259525299072266,
"learning_rate": 4.886720194242641e-05,
"loss": 1.3282,
"step": 4010
},
{
"epoch": 0.19,
"grad_norm": 0.40386712551116943,
"learning_rate": 4.886158555530373e-05,
"loss": 1.388,
"step": 4020
},
{
"epoch": 0.19,
"grad_norm": 0.3765401244163513,
"learning_rate": 4.8855955603995794e-05,
"loss": 1.3227,
"step": 4030
},
{
"epoch": 0.19,
"grad_norm": 0.5956104397773743,
"learning_rate": 4.885031209170299e-05,
"loss": 1.3523,
"step": 4040
},
{
"epoch": 0.19,
"grad_norm": 0.3925481140613556,
"learning_rate": 4.884465502163337e-05,
"loss": 1.3852,
"step": 4050
},
{
"epoch": 0.19,
"grad_norm": 0.3959333300590515,
"learning_rate": 4.883898439700272e-05,
"loss": 1.4371,
"step": 4060
},
{
"epoch": 0.2,
"grad_norm": 0.4429927468299866,
"learning_rate": 4.883330022103452e-05,
"loss": 1.3473,
"step": 4070
},
{
"epoch": 0.2,
"grad_norm": 0.3472846746444702,
"learning_rate": 4.8827602496959955e-05,
"loss": 1.3263,
"step": 4080
},
{
"epoch": 0.2,
"grad_norm": 0.396863728761673,
"learning_rate": 4.882189122801791e-05,
"loss": 1.3798,
"step": 4090
},
{
"epoch": 0.2,
"grad_norm": 0.41071972250938416,
"learning_rate": 4.8816166417454986e-05,
"loss": 1.4608,
"step": 4100
},
{
"epoch": 0.2,
"grad_norm": 0.33676210045814514,
"learning_rate": 4.881042806852546e-05,
"loss": 1.4436,
"step": 4110
},
{
"epoch": 0.2,
"grad_norm": 0.4066254496574402,
"learning_rate": 4.88046761844913e-05,
"loss": 1.3834,
"step": 4120
},
{
"epoch": 0.2,
"grad_norm": 0.6033533811569214,
"learning_rate": 4.87989107686222e-05,
"loss": 1.36,
"step": 4130
},
{
"epoch": 0.2,
"grad_norm": 0.3646985590457916,
"learning_rate": 4.8793131824195506e-05,
"loss": 1.4214,
"step": 4140
},
{
"epoch": 0.2,
"grad_norm": 0.3660110533237457,
"learning_rate": 4.87873393544963e-05,
"loss": 1.4369,
"step": 4150
},
{
"epoch": 0.2,
"grad_norm": 0.33257168531417847,
"learning_rate": 4.87815333628173e-05,
"loss": 1.3382,
"step": 4160
},
{
"epoch": 0.2,
"grad_norm": 0.33120933175086975,
"learning_rate": 4.877571385245896e-05,
"loss": 1.3748,
"step": 4170
},
{
"epoch": 0.2,
"grad_norm": 0.5946853160858154,
"learning_rate": 4.8769880826729385e-05,
"loss": 1.4279,
"step": 4180
},
{
"epoch": 0.2,
"grad_norm": 0.48969128727912903,
"learning_rate": 4.876403428894438e-05,
"loss": 1.3701,
"step": 4190
},
{
"epoch": 0.2,
"grad_norm": 0.3290250599384308,
"learning_rate": 4.8758174242427435e-05,
"loss": 1.4159,
"step": 4200
},
{
"epoch": 0.2,
"grad_norm": 0.46194833517074585,
"learning_rate": 4.875230069050969e-05,
"loss": 1.3957,
"step": 4210
},
{
"epoch": 0.2,
"grad_norm": 0.4537023603916168,
"learning_rate": 4.874641363653e-05,
"loss": 1.3963,
"step": 4220
},
{
"epoch": 0.2,
"grad_norm": 0.4122201204299927,
"learning_rate": 4.874051308383486e-05,
"loss": 1.373,
"step": 4230
},
{
"epoch": 0.2,
"grad_norm": 0.4980943202972412,
"learning_rate": 4.8734599035778465e-05,
"loss": 1.3668,
"step": 4240
},
{
"epoch": 0.2,
"grad_norm": 0.37470829486846924,
"learning_rate": 4.872867149572268e-05,
"loss": 1.4125,
"step": 4250
},
{
"epoch": 0.2,
"grad_norm": 0.3890114426612854,
"learning_rate": 4.872332517679774e-05,
"loss": 1.3383,
"step": 4260
},
{
"epoch": 0.2,
"grad_norm": 0.406592458486557,
"learning_rate": 4.8717372011232476e-05,
"loss": 1.3958,
"step": 4270
},
{
"epoch": 0.21,
"grad_norm": 0.392891526222229,
"learning_rate": 4.8711405363460573e-05,
"loss": 1.4101,
"step": 4280
},
{
"epoch": 0.21,
"grad_norm": 0.45984283089637756,
"learning_rate": 4.870542523687378e-05,
"loss": 1.3245,
"step": 4290
},
{
"epoch": 0.21,
"grad_norm": 0.5970633029937744,
"learning_rate": 4.8699431634871504e-05,
"loss": 1.3827,
"step": 4300
},
{
"epoch": 0.21,
"grad_norm": 0.39968982338905334,
"learning_rate": 4.869342456086084e-05,
"loss": 1.4472,
"step": 4310
},
{
"epoch": 0.21,
"grad_norm": 0.37416696548461914,
"learning_rate": 4.868740401825654e-05,
"loss": 1.3556,
"step": 4320
},
{
"epoch": 0.21,
"grad_norm": 0.3985888659954071,
"learning_rate": 4.868137001048098e-05,
"loss": 1.4256,
"step": 4330
},
{
"epoch": 0.21,
"grad_norm": 0.40038734674453735,
"learning_rate": 4.86753225409642e-05,
"loss": 1.4488,
"step": 4340
},
{
"epoch": 0.21,
"grad_norm": 0.38625702261924744,
"learning_rate": 4.866926161314391e-05,
"loss": 1.305,
"step": 4350
},
{
"epoch": 0.21,
"grad_norm": 0.39062339067459106,
"learning_rate": 4.866318723046547e-05,
"loss": 1.343,
"step": 4360
},
{
"epoch": 0.21,
"grad_norm": 0.3100876212120056,
"learning_rate": 4.865709939638188e-05,
"loss": 1.3523,
"step": 4370
},
{
"epoch": 0.21,
"grad_norm": 0.5294597148895264,
"learning_rate": 4.865099811435377e-05,
"loss": 1.2806,
"step": 4380
},
{
"epoch": 0.21,
"grad_norm": 0.3181302547454834,
"learning_rate": 4.864488338784944e-05,
"loss": 1.3895,
"step": 4390
},
{
"epoch": 0.21,
"grad_norm": 0.44422975182533264,
"learning_rate": 4.863875522034483e-05,
"loss": 1.4095,
"step": 4400
},
{
"epoch": 0.21,
"grad_norm": 0.3645181357860565,
"learning_rate": 4.86326136153235e-05,
"loss": 1.3349,
"step": 4410
},
{
"epoch": 0.21,
"grad_norm": 0.38209041953086853,
"learning_rate": 4.862645857627668e-05,
"loss": 1.4561,
"step": 4420
},
{
"epoch": 0.21,
"grad_norm": 0.3653987944126129,
"learning_rate": 4.8620290106703206e-05,
"loss": 1.4522,
"step": 4430
},
{
"epoch": 0.21,
"grad_norm": 0.4010675549507141,
"learning_rate": 4.8614108210109566e-05,
"loss": 1.4407,
"step": 4440
},
{
"epoch": 0.21,
"grad_norm": 0.4154636859893799,
"learning_rate": 4.860791289000988e-05,
"loss": 1.4498,
"step": 4450
},
{
"epoch": 0.21,
"grad_norm": 0.38323020935058594,
"learning_rate": 4.860170414992588e-05,
"loss": 1.4001,
"step": 4460
},
{
"epoch": 0.21,
"grad_norm": 0.35228949785232544,
"learning_rate": 4.859548199338696e-05,
"loss": 1.2732,
"step": 4470
},
{
"epoch": 0.22,
"grad_norm": 0.3578246533870697,
"learning_rate": 4.8589246423930115e-05,
"loss": 1.3831,
"step": 4480
},
{
"epoch": 0.22,
"grad_norm": 0.3850768506526947,
"learning_rate": 4.858299744509997e-05,
"loss": 1.405,
"step": 4490
},
{
"epoch": 0.22,
"grad_norm": 0.42587488889694214,
"learning_rate": 4.857673506044878e-05,
"loss": 1.3728,
"step": 4500
},
{
"epoch": 0.22,
"grad_norm": 0.4556770920753479,
"learning_rate": 4.857045927353641e-05,
"loss": 1.3961,
"step": 4510
},
{
"epoch": 0.22,
"grad_norm": 0.32983559370040894,
"learning_rate": 4.856417008793035e-05,
"loss": 1.4398,
"step": 4520
},
{
"epoch": 0.22,
"grad_norm": 0.4086816608905792,
"learning_rate": 4.8557867507205715e-05,
"loss": 1.49,
"step": 4530
},
{
"epoch": 0.22,
"grad_norm": 0.4496689736843109,
"learning_rate": 4.855155153494521e-05,
"loss": 1.4307,
"step": 4540
},
{
"epoch": 0.22,
"grad_norm": 0.45416951179504395,
"learning_rate": 4.854522217473916e-05,
"loss": 1.4282,
"step": 4550
},
{
"epoch": 0.22,
"grad_norm": 0.37094685435295105,
"learning_rate": 4.8538879430185536e-05,
"loss": 1.3778,
"step": 4560
},
{
"epoch": 0.22,
"grad_norm": 0.42879119515419006,
"learning_rate": 4.853252330488987e-05,
"loss": 1.2649,
"step": 4570
},
{
"epoch": 0.22,
"grad_norm": 0.3889993727207184,
"learning_rate": 4.8526153802465314e-05,
"loss": 1.4214,
"step": 4580
},
{
"epoch": 0.22,
"grad_norm": 0.3959541618824005,
"learning_rate": 4.851977092653264e-05,
"loss": 1.3704,
"step": 4590
},
{
"epoch": 0.22,
"grad_norm": 0.4111979901790619,
"learning_rate": 4.8513374680720214e-05,
"loss": 1.4188,
"step": 4600
},
{
"epoch": 0.22,
"grad_norm": 0.4254642724990845,
"learning_rate": 4.850696506866399e-05,
"loss": 1.4655,
"step": 4610
},
{
"epoch": 0.22,
"grad_norm": 0.4297488033771515,
"learning_rate": 4.850054209400755e-05,
"loss": 1.4849,
"step": 4620
},
{
"epoch": 0.22,
"grad_norm": 0.4221087694168091,
"learning_rate": 4.849410576040202e-05,
"loss": 1.4046,
"step": 4630
},
{
"epoch": 0.22,
"grad_norm": 0.402934730052948,
"learning_rate": 4.848765607150618e-05,
"loss": 1.3493,
"step": 4640
},
{
"epoch": 0.22,
"grad_norm": 0.40379470586776733,
"learning_rate": 4.848119303098636e-05,
"loss": 1.3663,
"step": 4650
},
{
"epoch": 0.22,
"grad_norm": 0.3863302171230316,
"learning_rate": 4.847471664251649e-05,
"loss": 1.3791,
"step": 4660
},
{
"epoch": 0.22,
"grad_norm": 0.36187705397605896,
"learning_rate": 4.8468226909778115e-05,
"loss": 1.4349,
"step": 4670
},
{
"epoch": 0.22,
"grad_norm": 0.3691658079624176,
"learning_rate": 4.846172383646031e-05,
"loss": 1.357,
"step": 4680
},
{
"epoch": 0.23,
"grad_norm": 0.3678319454193115,
"learning_rate": 4.8455207426259785e-05,
"loss": 1.33,
"step": 4690
},
{
"epoch": 0.23,
"grad_norm": 0.38880303502082825,
"learning_rate": 4.84486776828808e-05,
"loss": 1.3715,
"step": 4700
},
{
"epoch": 0.23,
"grad_norm": 0.4265637993812561,
"learning_rate": 4.844213461003522e-05,
"loss": 1.3235,
"step": 4710
},
{
"epoch": 0.23,
"grad_norm": 0.42313674092292786,
"learning_rate": 4.8435578211442464e-05,
"loss": 1.4205,
"step": 4720
},
{
"epoch": 0.23,
"grad_norm": 0.37444865703582764,
"learning_rate": 4.842900849082953e-05,
"loss": 1.418,
"step": 4730
},
{
"epoch": 0.23,
"grad_norm": 0.3745338022708893,
"learning_rate": 4.842242545193101e-05,
"loss": 1.3994,
"step": 4740
},
{
"epoch": 0.23,
"grad_norm": 0.4207676649093628,
"learning_rate": 4.841582909848904e-05,
"loss": 1.4339,
"step": 4750
},
{
"epoch": 0.23,
"grad_norm": 0.41867244243621826,
"learning_rate": 4.840921943425333e-05,
"loss": 1.3798,
"step": 4760
},
{
"epoch": 0.23,
"grad_norm": 0.4031929075717926,
"learning_rate": 4.8402596462981174e-05,
"loss": 1.4857,
"step": 4770
},
{
"epoch": 0.23,
"grad_norm": 0.35046666860580444,
"learning_rate": 4.839596018843742e-05,
"loss": 1.3885,
"step": 4780
},
{
"epoch": 0.23,
"grad_norm": 0.4085848033428192,
"learning_rate": 4.8389310614394456e-05,
"loss": 1.385,
"step": 4790
},
{
"epoch": 0.23,
"grad_norm": 0.38278183341026306,
"learning_rate": 4.838264774463228e-05,
"loss": 1.3926,
"step": 4800
},
{
"epoch": 0.23,
"grad_norm": 0.37496405839920044,
"learning_rate": 4.8375971582938394e-05,
"loss": 1.3804,
"step": 4810
},
{
"epoch": 0.23,
"grad_norm": 0.3691863715648651,
"learning_rate": 4.836928213310791e-05,
"loss": 1.4247,
"step": 4820
},
{
"epoch": 0.23,
"grad_norm": 0.3793914318084717,
"learning_rate": 4.836257939894343e-05,
"loss": 1.37,
"step": 4830
},
{
"epoch": 0.23,
"grad_norm": 0.3822409510612488,
"learning_rate": 4.835586338425516e-05,
"loss": 1.4453,
"step": 4840
},
{
"epoch": 0.23,
"grad_norm": 0.4485887885093689,
"learning_rate": 4.834913409286084e-05,
"loss": 1.3788,
"step": 4850
},
{
"epoch": 0.23,
"grad_norm": 0.40386226773262024,
"learning_rate": 4.834239152858575e-05,
"loss": 1.5042,
"step": 4860
},
{
"epoch": 0.23,
"grad_norm": 0.4584513306617737,
"learning_rate": 4.8335635695262716e-05,
"loss": 1.3723,
"step": 4870
},
{
"epoch": 0.23,
"grad_norm": 0.606304943561554,
"learning_rate": 4.8328866596732114e-05,
"loss": 1.4122,
"step": 4880
},
{
"epoch": 0.23,
"grad_norm": 0.5519858002662659,
"learning_rate": 4.832208423684186e-05,
"loss": 1.2852,
"step": 4890
},
{
"epoch": 0.24,
"grad_norm": 0.3566693365573883,
"learning_rate": 4.831528861944741e-05,
"loss": 1.3997,
"step": 4900
},
{
"epoch": 0.24,
"grad_norm": 0.37323299050331116,
"learning_rate": 4.830847974841174e-05,
"loss": 1.3306,
"step": 4910
},
{
"epoch": 0.24,
"grad_norm": 0.3981427848339081,
"learning_rate": 4.830165762760537e-05,
"loss": 1.3239,
"step": 4920
},
{
"epoch": 0.24,
"grad_norm": 0.40320611000061035,
"learning_rate": 4.8294822260906365e-05,
"loss": 1.3991,
"step": 4930
},
{
"epoch": 0.24,
"grad_norm": 0.3638981878757477,
"learning_rate": 4.82879736522003e-05,
"loss": 1.3288,
"step": 4940
},
{
"epoch": 0.24,
"grad_norm": 0.5388992428779602,
"learning_rate": 4.8281111805380296e-05,
"loss": 1.4235,
"step": 4950
},
{
"epoch": 0.24,
"grad_norm": 0.4435693025588989,
"learning_rate": 4.827423672434698e-05,
"loss": 1.4899,
"step": 4960
},
{
"epoch": 0.24,
"grad_norm": 0.4411034882068634,
"learning_rate": 4.826734841300851e-05,
"loss": 1.4055,
"step": 4970
},
{
"epoch": 0.24,
"grad_norm": 0.3863956034183502,
"learning_rate": 4.826044687528058e-05,
"loss": 1.3323,
"step": 4980
},
{
"epoch": 0.24,
"grad_norm": 0.3567318320274353,
"learning_rate": 4.825353211508636e-05,
"loss": 1.4927,
"step": 4990
},
{
"epoch": 0.24,
"grad_norm": 0.4811917245388031,
"learning_rate": 4.82466041363566e-05,
"loss": 1.4119,
"step": 5000
},
{
"epoch": 0.24,
"grad_norm": 0.4191003441810608,
"learning_rate": 4.82396629430295e-05,
"loss": 1.351,
"step": 5010
},
{
"epoch": 0.24,
"grad_norm": 0.37054839730262756,
"learning_rate": 4.823270853905082e-05,
"loss": 1.3923,
"step": 5020
},
{
"epoch": 0.24,
"grad_norm": 0.36948150396347046,
"learning_rate": 4.8225740928373796e-05,
"loss": 1.3282,
"step": 5030
},
{
"epoch": 0.24,
"grad_norm": 0.39264991879463196,
"learning_rate": 4.8218760114959193e-05,
"loss": 1.3958,
"step": 5040
},
{
"epoch": 0.24,
"grad_norm": 0.406788170337677,
"learning_rate": 4.821176610277528e-05,
"loss": 1.4312,
"step": 5050
},
{
"epoch": 0.24,
"grad_norm": 0.3900972306728363,
"learning_rate": 4.820475889579782e-05,
"loss": 1.3631,
"step": 5060
},
{
"epoch": 0.24,
"grad_norm": 0.3968881070613861,
"learning_rate": 4.819773849801007e-05,
"loss": 1.3002,
"step": 5070
},
{
"epoch": 0.24,
"grad_norm": 0.3620697557926178,
"learning_rate": 4.8190704913402793e-05,
"loss": 1.4096,
"step": 5080
},
{
"epoch": 0.24,
"grad_norm": 0.41546300053596497,
"learning_rate": 4.818365814597427e-05,
"loss": 1.4787,
"step": 5090
},
{
"epoch": 0.24,
"grad_norm": 0.33438408374786377,
"learning_rate": 4.817659819973024e-05,
"loss": 1.3449,
"step": 5100
},
{
"epoch": 0.25,
"grad_norm": 0.47801473736763,
"learning_rate": 4.8169525078683955e-05,
"loss": 1.346,
"step": 5110
},
{
"epoch": 0.25,
"grad_norm": 0.49584850668907166,
"learning_rate": 4.816243878685614e-05,
"loss": 1.4183,
"step": 5120
},
{
"epoch": 0.25,
"grad_norm": 0.34061333537101746,
"learning_rate": 4.815533932827503e-05,
"loss": 1.3909,
"step": 5130
},
{
"epoch": 0.25,
"grad_norm": 0.30921533703804016,
"learning_rate": 4.814822670697633e-05,
"loss": 1.363,
"step": 5140
},
{
"epoch": 0.25,
"grad_norm": 0.3006262183189392,
"learning_rate": 4.8141100927003224e-05,
"loss": 1.4285,
"step": 5150
},
{
"epoch": 0.25,
"grad_norm": 0.3685610294342041,
"learning_rate": 4.813396199240639e-05,
"loss": 1.425,
"step": 5160
},
{
"epoch": 0.25,
"grad_norm": 0.5372167229652405,
"learning_rate": 4.812680990724397e-05,
"loss": 1.394,
"step": 5170
},
{
"epoch": 0.25,
"grad_norm": 0.35630884766578674,
"learning_rate": 4.811964467558158e-05,
"loss": 1.4641,
"step": 5180
},
{
"epoch": 0.25,
"grad_norm": 0.39942196011543274,
"learning_rate": 4.811246630149233e-05,
"loss": 1.3976,
"step": 5190
},
{
"epoch": 0.25,
"grad_norm": 0.44498565793037415,
"learning_rate": 4.810527478905678e-05,
"loss": 1.3321,
"step": 5200
},
{
"epoch": 0.25,
"grad_norm": 0.4407236874103546,
"learning_rate": 4.809807014236296e-05,
"loss": 1.3932,
"step": 5210
},
{
"epoch": 0.25,
"grad_norm": 0.4077320098876953,
"learning_rate": 4.809085236550639e-05,
"loss": 1.4089,
"step": 5220
},
{
"epoch": 0.25,
"grad_norm": 0.3109973669052124,
"learning_rate": 4.8083621462590014e-05,
"loss": 1.3623,
"step": 5230
},
{
"epoch": 0.25,
"grad_norm": 0.299821138381958,
"learning_rate": 4.807637743772426e-05,
"loss": 1.369,
"step": 5240
},
{
"epoch": 0.25,
"grad_norm": 0.3847900629043579,
"learning_rate": 4.806912029502703e-05,
"loss": 1.3162,
"step": 5250
},
{
"epoch": 0.25,
"grad_norm": 0.7119512557983398,
"learning_rate": 4.8061850038623664e-05,
"loss": 1.339,
"step": 5260
},
{
"epoch": 0.25,
"grad_norm": 0.4665127694606781,
"learning_rate": 4.805529559905748e-05,
"loss": 1.4218,
"step": 5270
},
{
"epoch": 0.25,
"grad_norm": 0.4741785526275635,
"learning_rate": 4.804800043800446e-05,
"loss": 1.4232,
"step": 5280
},
{
"epoch": 0.25,
"grad_norm": 0.4746471643447876,
"learning_rate": 4.804142359098447e-05,
"loss": 1.4081,
"step": 5290
},
{
"epoch": 0.25,
"grad_norm": 0.3058326840400696,
"learning_rate": 4.803410354025231e-05,
"loss": 1.3856,
"step": 5300
},
{
"epoch": 0.25,
"grad_norm": 0.35912251472473145,
"learning_rate": 4.802677039571937e-05,
"loss": 1.4245,
"step": 5310
},
{
"epoch": 0.26,
"grad_norm": 0.4378702938556671,
"learning_rate": 4.80194241615542e-05,
"loss": 1.3573,
"step": 5320
},
{
"epoch": 0.26,
"grad_norm": 0.3681640326976776,
"learning_rate": 4.8012064841932815e-05,
"loss": 1.3586,
"step": 5330
},
{
"epoch": 0.26,
"grad_norm": 0.4573310315608978,
"learning_rate": 4.800469244103861e-05,
"loss": 1.3467,
"step": 5340
},
{
"epoch": 0.26,
"grad_norm": 0.406074583530426,
"learning_rate": 4.7997306963062466e-05,
"loss": 1.3036,
"step": 5350
},
{
"epoch": 0.26,
"grad_norm": 0.47218167781829834,
"learning_rate": 4.798990841220267e-05,
"loss": 1.2858,
"step": 5360
},
{
"epoch": 0.26,
"grad_norm": 0.4285829961299896,
"learning_rate": 4.7982496792664944e-05,
"loss": 1.429,
"step": 5370
},
{
"epoch": 0.26,
"grad_norm": 0.40524205565452576,
"learning_rate": 4.797507210866245e-05,
"loss": 1.322,
"step": 5380
},
{
"epoch": 0.26,
"grad_norm": 0.36471444368362427,
"learning_rate": 4.7967634364415784e-05,
"loss": 1.3682,
"step": 5390
},
{
"epoch": 0.26,
"grad_norm": 0.5011775493621826,
"learning_rate": 4.796018356415293e-05,
"loss": 1.4415,
"step": 5400
},
{
"epoch": 0.26,
"grad_norm": 0.37677791714668274,
"learning_rate": 4.795271971210932e-05,
"loss": 1.3064,
"step": 5410
},
{
"epoch": 0.26,
"grad_norm": 0.3738216459751129,
"learning_rate": 4.794524281252781e-05,
"loss": 1.3535,
"step": 5420
},
{
"epoch": 0.26,
"grad_norm": 0.36507245898246765,
"learning_rate": 4.793775286965867e-05,
"loss": 1.4163,
"step": 5430
},
{
"epoch": 0.26,
"grad_norm": 0.4252508878707886,
"learning_rate": 4.7930249887759556e-05,
"loss": 1.3364,
"step": 5440
},
{
"epoch": 0.26,
"grad_norm": 0.38034453988075256,
"learning_rate": 4.7922733871095586e-05,
"loss": 1.4196,
"step": 5450
},
{
"epoch": 0.26,
"grad_norm": 0.3702666163444519,
"learning_rate": 4.791520482393924e-05,
"loss": 1.357,
"step": 5460
},
{
"epoch": 0.26,
"grad_norm": 0.37074029445648193,
"learning_rate": 4.790766275057044e-05,
"loss": 1.4188,
"step": 5470
},
{
"epoch": 0.26,
"grad_norm": 0.3444986045360565,
"learning_rate": 4.79001076552765e-05,
"loss": 1.4278,
"step": 5480
},
{
"epoch": 0.26,
"grad_norm": 0.44791311025619507,
"learning_rate": 4.789253954235211e-05,
"loss": 1.3615,
"step": 5490
},
{
"epoch": 0.26,
"grad_norm": 0.3411417305469513,
"learning_rate": 4.788495841609942e-05,
"loss": 1.4096,
"step": 5500
},
{
"epoch": 0.26,
"grad_norm": 0.4716344475746155,
"learning_rate": 4.787736428082792e-05,
"loss": 1.2935,
"step": 5510
},
{
"epoch": 0.26,
"grad_norm": 0.41569599509239197,
"learning_rate": 4.786975714085452e-05,
"loss": 1.4494,
"step": 5520
},
{
"epoch": 0.27,
"grad_norm": 0.5307323336601257,
"learning_rate": 4.786213700050353e-05,
"loss": 1.3822,
"step": 5530
},
{
"epoch": 0.27,
"grad_norm": 0.3211369514465332,
"learning_rate": 4.7854503864106626e-05,
"loss": 1.3261,
"step": 5540
},
{
"epoch": 0.27,
"grad_norm": 0.4213569164276123,
"learning_rate": 4.784685773600289e-05,
"loss": 1.312,
"step": 5550
},
{
"epoch": 0.27,
"grad_norm": 0.36504948139190674,
"learning_rate": 4.78391986205388e-05,
"loss": 1.4143,
"step": 5560
},
{
"epoch": 0.27,
"grad_norm": 0.4195592701435089,
"learning_rate": 4.783152652206818e-05,
"loss": 1.4465,
"step": 5570
},
{
"epoch": 0.27,
"grad_norm": 0.380455881357193,
"learning_rate": 4.7823841444952275e-05,
"loss": 1.3648,
"step": 5580
},
{
"epoch": 0.27,
"grad_norm": 0.40087655186653137,
"learning_rate": 4.781614339355968e-05,
"loss": 1.3688,
"step": 5590
},
{
"epoch": 0.27,
"grad_norm": 0.4018709361553192,
"learning_rate": 4.780843237226637e-05,
"loss": 1.4285,
"step": 5600
},
{
"epoch": 0.27,
"grad_norm": 0.3394882380962372,
"learning_rate": 4.780070838545569e-05,
"loss": 1.3086,
"step": 5610
},
{
"epoch": 0.27,
"grad_norm": 0.4284053444862366,
"learning_rate": 4.7792971437518384e-05,
"loss": 1.4391,
"step": 5620
},
{
"epoch": 0.27,
"grad_norm": 0.3496474027633667,
"learning_rate": 4.778522153285253e-05,
"loss": 1.3743,
"step": 5630
},
{
"epoch": 0.27,
"grad_norm": 0.4204890727996826,
"learning_rate": 4.7777458675863573e-05,
"loss": 1.4579,
"step": 5640
},
{
"epoch": 0.27,
"grad_norm": 0.39384838938713074,
"learning_rate": 4.776968287096435e-05,
"loss": 1.3294,
"step": 5650
},
{
"epoch": 0.27,
"grad_norm": 0.36904141306877136,
"learning_rate": 4.776189412257502e-05,
"loss": 1.3755,
"step": 5660
},
{
"epoch": 0.27,
"grad_norm": 0.5375967025756836,
"learning_rate": 4.7754092435123145e-05,
"loss": 1.3835,
"step": 5670
},
{
"epoch": 0.27,
"grad_norm": 0.6000000834465027,
"learning_rate": 4.7746277813043586e-05,
"loss": 1.3311,
"step": 5680
},
{
"epoch": 0.27,
"grad_norm": 0.40945225954055786,
"learning_rate": 4.773845026077861e-05,
"loss": 1.3532,
"step": 5690
},
{
"epoch": 0.27,
"grad_norm": 0.6537255048751831,
"learning_rate": 4.773060978277779e-05,
"loss": 1.3308,
"step": 5700
},
{
"epoch": 0.27,
"grad_norm": 0.363484650850296,
"learning_rate": 4.772275638349809e-05,
"loss": 1.3619,
"step": 5710
},
{
"epoch": 0.27,
"grad_norm": 0.35903969407081604,
"learning_rate": 4.7714890067403784e-05,
"loss": 1.4389,
"step": 5720
},
{
"epoch": 0.28,
"grad_norm": 0.30839696526527405,
"learning_rate": 4.77070108389665e-05,
"loss": 1.3825,
"step": 5730
},
{
"epoch": 0.28,
"grad_norm": 0.43783774971961975,
"learning_rate": 4.7699118702665215e-05,
"loss": 1.35,
"step": 5740
},
{
"epoch": 0.28,
"grad_norm": 0.39843234419822693,
"learning_rate": 4.7691213662986224e-05,
"loss": 1.4307,
"step": 5750
},
{
"epoch": 0.28,
"grad_norm": 0.4051912724971771,
"learning_rate": 4.7683295724423184e-05,
"loss": 1.3154,
"step": 5760
},
{
"epoch": 0.28,
"grad_norm": 0.3443717956542969,
"learning_rate": 4.767536489147707e-05,
"loss": 1.3667,
"step": 5770
},
{
"epoch": 0.28,
"grad_norm": 0.32003670930862427,
"learning_rate": 4.766742116865617e-05,
"loss": 1.4204,
"step": 5780
},
{
"epoch": 0.28,
"grad_norm": 0.31230783462524414,
"learning_rate": 4.765946456047612e-05,
"loss": 1.4103,
"step": 5790
},
{
"epoch": 0.28,
"grad_norm": 0.414163202047348,
"learning_rate": 4.765149507145988e-05,
"loss": 1.4451,
"step": 5800
},
{
"epoch": 0.28,
"grad_norm": 0.42239320278167725,
"learning_rate": 4.7643512706137736e-05,
"loss": 1.3554,
"step": 5810
},
{
"epoch": 0.28,
"grad_norm": 0.41019243001937866,
"learning_rate": 4.763551746904728e-05,
"loss": 1.4001,
"step": 5820
},
{
"epoch": 0.28,
"grad_norm": 0.38920360803604126,
"learning_rate": 4.7627509364733426e-05,
"loss": 1.4065,
"step": 5830
},
{
"epoch": 0.28,
"grad_norm": 0.396381676197052,
"learning_rate": 4.7619488397748404e-05,
"loss": 1.3703,
"step": 5840
},
{
"epoch": 0.28,
"grad_norm": 0.35331448912620544,
"learning_rate": 4.7611454572651746e-05,
"loss": 1.4084,
"step": 5850
},
{
"epoch": 0.28,
"grad_norm": 0.3805950880050659,
"learning_rate": 4.7603407894010314e-05,
"loss": 1.4661,
"step": 5860
},
{
"epoch": 0.28,
"grad_norm": 0.4126935601234436,
"learning_rate": 4.759534836639826e-05,
"loss": 1.377,
"step": 5870
},
{
"epoch": 0.28,
"grad_norm": 0.37023454904556274,
"learning_rate": 4.758727599439705e-05,
"loss": 1.3895,
"step": 5880
},
{
"epoch": 0.28,
"grad_norm": 0.4315061867237091,
"learning_rate": 4.757919078259545e-05,
"loss": 1.3558,
"step": 5890
},
{
"epoch": 0.28,
"grad_norm": 0.35196399688720703,
"learning_rate": 4.757109273558951e-05,
"loss": 1.3634,
"step": 5900
},
{
"epoch": 0.28,
"grad_norm": 0.42198696732521057,
"learning_rate": 4.756298185798259e-05,
"loss": 1.4272,
"step": 5910
},
{
"epoch": 0.28,
"grad_norm": 0.3349405825138092,
"learning_rate": 4.755485815438534e-05,
"loss": 1.3878,
"step": 5920
},
{
"epoch": 0.28,
"grad_norm": 0.4049832820892334,
"learning_rate": 4.7546721629415717e-05,
"loss": 1.3562,
"step": 5930
},
{
"epoch": 0.29,
"grad_norm": 0.3526436984539032,
"learning_rate": 4.7538572287698924e-05,
"loss": 1.3119,
"step": 5940
},
{
"epoch": 0.29,
"grad_norm": 0.4919440746307373,
"learning_rate": 4.75304101338675e-05,
"loss": 1.4221,
"step": 5950
},
{
"epoch": 0.29,
"grad_norm": 0.41213706135749817,
"learning_rate": 4.7522235172561245e-05,
"loss": 1.4134,
"step": 5960
},
{
"epoch": 0.29,
"grad_norm": 0.3352992534637451,
"learning_rate": 4.751404740842722e-05,
"loss": 1.3312,
"step": 5970
},
{
"epoch": 0.29,
"grad_norm": 0.35339492559432983,
"learning_rate": 4.7505846846119806e-05,
"loss": 1.4077,
"step": 5980
},
{
"epoch": 0.29,
"grad_norm": 0.3903467059135437,
"learning_rate": 4.749763349030061e-05,
"loss": 1.3834,
"step": 5990
},
{
"epoch": 0.29,
"grad_norm": 0.3679284453392029,
"learning_rate": 4.748940734563856e-05,
"loss": 1.4557,
"step": 6000
},
{
"epoch": 0.29,
"grad_norm": 0.3598511815071106,
"learning_rate": 4.748116841680982e-05,
"loss": 1.3881,
"step": 6010
},
{
"epoch": 0.29,
"grad_norm": 0.351188600063324,
"learning_rate": 4.747291670849783e-05,
"loss": 1.4115,
"step": 6020
},
{
"epoch": 0.29,
"grad_norm": 0.4349496364593506,
"learning_rate": 4.7464652225393294e-05,
"loss": 1.3935,
"step": 6030
},
{
"epoch": 0.29,
"grad_norm": 0.4459682106971741,
"learning_rate": 4.745637497219419e-05,
"loss": 1.4308,
"step": 6040
},
{
"epoch": 0.29,
"grad_norm": 0.4155092239379883,
"learning_rate": 4.744808495360575e-05,
"loss": 1.3523,
"step": 6050
},
{
"epoch": 0.29,
"grad_norm": 0.440679132938385,
"learning_rate": 4.743978217434045e-05,
"loss": 1.3699,
"step": 6060
},
{
"epoch": 0.29,
"grad_norm": 0.40542155504226685,
"learning_rate": 4.7431466639118026e-05,
"loss": 1.3596,
"step": 6070
},
{
"epoch": 0.29,
"grad_norm": 0.39679834246635437,
"learning_rate": 4.7423138352665475e-05,
"loss": 1.4466,
"step": 6080
},
{
"epoch": 0.29,
"grad_norm": 0.35273149609565735,
"learning_rate": 4.741479731971703e-05,
"loss": 1.3883,
"step": 6090
},
{
"epoch": 0.29,
"grad_norm": 0.3129054307937622,
"learning_rate": 4.7406443545014166e-05,
"loss": 1.4003,
"step": 6100
},
{
"epoch": 0.29,
"grad_norm": 0.35774657130241394,
"learning_rate": 4.739807703330563e-05,
"loss": 1.3843,
"step": 6110
},
{
"epoch": 0.29,
"grad_norm": 0.4142649471759796,
"learning_rate": 4.7389697789347364e-05,
"loss": 1.4983,
"step": 6120
},
{
"epoch": 0.29,
"grad_norm": 0.4555107057094574,
"learning_rate": 4.73813058179026e-05,
"loss": 1.2857,
"step": 6130
},
{
"epoch": 0.29,
"grad_norm": 0.3330358564853668,
"learning_rate": 4.7372901123741754e-05,
"loss": 1.3275,
"step": 6140
},
{
"epoch": 0.3,
"grad_norm": 0.35886284708976746,
"learning_rate": 4.7364483711642514e-05,
"loss": 1.4174,
"step": 6150
},
{
"epoch": 0.3,
"grad_norm": 0.43242907524108887,
"learning_rate": 4.735605358638977e-05,
"loss": 1.3796,
"step": 6160
},
{
"epoch": 0.3,
"grad_norm": 0.3526054918766022,
"learning_rate": 4.7347610752775664e-05,
"loss": 1.475,
"step": 6170
},
{
"epoch": 0.3,
"grad_norm": 0.3490343987941742,
"learning_rate": 4.733915521559954e-05,
"loss": 1.3558,
"step": 6180
},
{
"epoch": 0.3,
"grad_norm": 0.43646037578582764,
"learning_rate": 4.733068697966796e-05,
"loss": 1.4273,
"step": 6190
},
{
"epoch": 0.3,
"grad_norm": 0.34763920307159424,
"learning_rate": 4.732220604979474e-05,
"loss": 1.4074,
"step": 6200
},
{
"epoch": 0.3,
"grad_norm": 0.3755379319190979,
"learning_rate": 4.731371243080087e-05,
"loss": 1.3552,
"step": 6210
},
{
"epoch": 0.3,
"grad_norm": 0.4469471871852875,
"learning_rate": 4.730520612751458e-05,
"loss": 1.3739,
"step": 6220
},
{
"epoch": 0.3,
"grad_norm": 0.3868820369243622,
"learning_rate": 4.72966871447713e-05,
"loss": 1.3759,
"step": 6230
},
{
"epoch": 0.3,
"grad_norm": 0.3981528580188751,
"learning_rate": 4.7288155487413674e-05,
"loss": 1.3119,
"step": 6240
},
{
"epoch": 0.3,
"grad_norm": 0.4281034469604492,
"learning_rate": 4.727961116029154e-05,
"loss": 1.3414,
"step": 6250
},
{
"epoch": 0.3,
"grad_norm": 0.34223461151123047,
"learning_rate": 4.727105416826195e-05,
"loss": 1.3775,
"step": 6260
},
{
"epoch": 0.3,
"grad_norm": 0.4001463055610657,
"learning_rate": 4.726248451618915e-05,
"loss": 1.3207,
"step": 6270
},
{
"epoch": 0.3,
"grad_norm": 0.41338250041007996,
"learning_rate": 4.7253902208944573e-05,
"loss": 1.371,
"step": 6280
},
{
"epoch": 0.3,
"grad_norm": 0.44054630398750305,
"learning_rate": 4.7245307251406865e-05,
"loss": 1.3232,
"step": 6290
},
{
"epoch": 0.3,
"grad_norm": 0.33765530586242676,
"learning_rate": 4.723669964846186e-05,
"loss": 1.4129,
"step": 6300
},
{
"epoch": 0.3,
"grad_norm": 0.3949221670627594,
"learning_rate": 4.722807940500255e-05,
"loss": 1.3147,
"step": 6310
},
{
"epoch": 0.3,
"grad_norm": 0.36308789253234863,
"learning_rate": 4.7219446525929164e-05,
"loss": 1.3553,
"step": 6320
},
{
"epoch": 0.3,
"grad_norm": 0.38808193802833557,
"learning_rate": 4.721080101614907e-05,
"loss": 1.3522,
"step": 6330
},
{
"epoch": 0.3,
"grad_norm": 0.501069188117981,
"learning_rate": 4.7202142880576836e-05,
"loss": 1.3742,
"step": 6340
},
{
"epoch": 0.3,
"grad_norm": 0.36833855509757996,
"learning_rate": 4.719433976757722e-05,
"loss": 1.4327,
"step": 6350
},
{
"epoch": 0.31,
"grad_norm": 0.41826024651527405,
"learning_rate": 4.718565765656525e-05,
"loss": 1.3225,
"step": 6360
},
{
"epoch": 0.31,
"grad_norm": 0.37621310353279114,
"learning_rate": 4.7176962934053956e-05,
"loss": 1.3502,
"step": 6370
},
{
"epoch": 0.31,
"grad_norm": 0.3487604856491089,
"learning_rate": 4.716825560498587e-05,
"loss": 1.3937,
"step": 6380
},
{
"epoch": 0.31,
"grad_norm": 0.3775159418582916,
"learning_rate": 4.715953567431071e-05,
"loss": 1.3062,
"step": 6390
},
{
"epoch": 0.31,
"grad_norm": 0.37354177236557007,
"learning_rate": 4.715080314698533e-05,
"loss": 1.3697,
"step": 6400
},
{
"epoch": 0.31,
"grad_norm": 0.34692302346229553,
"learning_rate": 4.714205802797378e-05,
"loss": 1.3289,
"step": 6410
},
{
"epoch": 0.31,
"grad_norm": 0.3825347125530243,
"learning_rate": 4.713330032224723e-05,
"loss": 1.3635,
"step": 6420
},
{
"epoch": 0.31,
"grad_norm": 0.3910905420780182,
"learning_rate": 4.712540762956647e-05,
"loss": 1.4054,
"step": 6430
},
{
"epoch": 0.31,
"grad_norm": 0.31619513034820557,
"learning_rate": 4.711662602280267e-05,
"loss": 1.3262,
"step": 6440
},
{
"epoch": 0.31,
"grad_norm": 0.5018675327301025,
"learning_rate": 4.710783184378077e-05,
"loss": 1.3868,
"step": 6450
},
{
"epoch": 0.31,
"grad_norm": 0.36038869619369507,
"learning_rate": 4.709902509749985e-05,
"loss": 1.3639,
"step": 6460
},
{
"epoch": 0.31,
"grad_norm": 0.4369955360889435,
"learning_rate": 4.709020578896613e-05,
"loss": 1.3192,
"step": 6470
},
{
"epoch": 0.31,
"grad_norm": 0.3579140305519104,
"learning_rate": 4.708137392319296e-05,
"loss": 1.3302,
"step": 6480
},
{
"epoch": 0.31,
"grad_norm": 0.6790784597396851,
"learning_rate": 4.7072529505200834e-05,
"loss": 1.3539,
"step": 6490
},
{
"epoch": 0.31,
"grad_norm": 0.3491171598434448,
"learning_rate": 4.706367254001741e-05,
"loss": 1.3861,
"step": 6500
},
{
"epoch": 0.31,
"grad_norm": 0.3110009729862213,
"learning_rate": 4.7054803032677444e-05,
"loss": 1.3774,
"step": 6510
},
{
"epoch": 0.31,
"grad_norm": 0.3939664661884308,
"learning_rate": 4.704592098822283e-05,
"loss": 1.349,
"step": 6520
},
{
"epoch": 0.31,
"grad_norm": 0.4585970342159271,
"learning_rate": 4.703702641170258e-05,
"loss": 1.4923,
"step": 6530
},
{
"epoch": 0.31,
"grad_norm": 0.39042964577674866,
"learning_rate": 4.702811930817286e-05,
"loss": 1.4683,
"step": 6540
},
{
"epoch": 0.31,
"grad_norm": 0.44418269395828247,
"learning_rate": 4.701919968269694e-05,
"loss": 1.4014,
"step": 6550
},
{
"epoch": 0.31,
"grad_norm": 0.4168784022331238,
"learning_rate": 4.7010267540345176e-05,
"loss": 1.3047,
"step": 6560
},
{
"epoch": 0.32,
"grad_norm": 0.4119689166545868,
"learning_rate": 4.7001322886195096e-05,
"loss": 1.3732,
"step": 6570
},
{
"epoch": 0.32,
"grad_norm": 0.3707830309867859,
"learning_rate": 4.699236572533132e-05,
"loss": 1.3165,
"step": 6580
},
{
"epoch": 0.32,
"grad_norm": 0.6226394176483154,
"learning_rate": 4.698339606284554e-05,
"loss": 1.4104,
"step": 6590
},
{
"epoch": 0.32,
"grad_norm": 0.3444110155105591,
"learning_rate": 4.697441390383662e-05,
"loss": 1.3582,
"step": 6600
},
{
"epoch": 0.32,
"grad_norm": 0.3979819715023041,
"learning_rate": 4.696541925341047e-05,
"loss": 1.3075,
"step": 6610
},
{
"epoch": 0.32,
"grad_norm": 0.44674932956695557,
"learning_rate": 4.695641211668014e-05,
"loss": 1.4297,
"step": 6620
},
{
"epoch": 0.32,
"grad_norm": 0.3688850998878479,
"learning_rate": 4.6947392498765744e-05,
"loss": 1.402,
"step": 6630
},
{
"epoch": 0.32,
"grad_norm": 0.3981863558292389,
"learning_rate": 4.6938360404794526e-05,
"loss": 1.5001,
"step": 6640
},
{
"epoch": 0.32,
"grad_norm": 0.3663766384124756,
"learning_rate": 4.692931583990079e-05,
"loss": 1.3796,
"step": 6650
},
{
"epoch": 0.32,
"grad_norm": 0.33633315563201904,
"learning_rate": 4.692025880922596e-05,
"loss": 1.3343,
"step": 6660
},
{
"epoch": 0.32,
"grad_norm": 0.4531651735305786,
"learning_rate": 4.6911189317918525e-05,
"loss": 1.3604,
"step": 6670
},
{
"epoch": 0.32,
"grad_norm": 0.43992722034454346,
"learning_rate": 4.690210737113406e-05,
"loss": 1.4471,
"step": 6680
},
{
"epoch": 0.32,
"grad_norm": 0.32002195715904236,
"learning_rate": 4.6893012974035223e-05,
"loss": 1.3777,
"step": 6690
},
{
"epoch": 0.32,
"grad_norm": 0.5869015455245972,
"learning_rate": 4.688390613179174e-05,
"loss": 1.4541,
"step": 6700
},
{
"epoch": 0.32,
"grad_norm": 0.3732210099697113,
"learning_rate": 4.687478684958045e-05,
"loss": 1.3643,
"step": 6710
},
{
"epoch": 0.32,
"grad_norm": 0.3989577293395996,
"learning_rate": 4.6865655132585204e-05,
"loss": 1.3675,
"step": 6720
},
{
"epoch": 0.32,
"grad_norm": 0.3631729185581207,
"learning_rate": 4.685651098599697e-05,
"loss": 1.3499,
"step": 6730
},
{
"epoch": 0.32,
"grad_norm": 0.44370928406715393,
"learning_rate": 4.684735441501376e-05,
"loss": 1.4344,
"step": 6740
},
{
"epoch": 0.32,
"grad_norm": 0.5211467146873474,
"learning_rate": 4.683818542484065e-05,
"loss": 1.2788,
"step": 6750
},
{
"epoch": 0.32,
"grad_norm": 0.3690240979194641,
"learning_rate": 4.6829004020689775e-05,
"loss": 1.4085,
"step": 6760
},
{
"epoch": 0.32,
"grad_norm": 0.4505044221878052,
"learning_rate": 4.6819810207780336e-05,
"loss": 1.4283,
"step": 6770
},
{
"epoch": 0.33,
"grad_norm": 0.3633420169353485,
"learning_rate": 4.681060399133857e-05,
"loss": 1.3331,
"step": 6780
},
{
"epoch": 0.33,
"grad_norm": 0.3736770749092102,
"learning_rate": 4.680138537659781e-05,
"loss": 1.4595,
"step": 6790
},
{
"epoch": 0.33,
"grad_norm": 0.5162922143936157,
"learning_rate": 4.679215436879836e-05,
"loss": 1.3106,
"step": 6800
},
{
"epoch": 0.33,
"grad_norm": 0.3515087068080902,
"learning_rate": 4.678291097318763e-05,
"loss": 1.3367,
"step": 6810
},
{
"epoch": 0.33,
"grad_norm": 0.5579962134361267,
"learning_rate": 4.677365519502007e-05,
"loss": 1.4005,
"step": 6820
},
{
"epoch": 0.33,
"grad_norm": 0.39628589153289795,
"learning_rate": 4.676438703955714e-05,
"loss": 1.4726,
"step": 6830
},
{
"epoch": 0.33,
"grad_norm": 0.3891077935695648,
"learning_rate": 4.675510651206732e-05,
"loss": 1.3936,
"step": 6840
},
{
"epoch": 0.33,
"grad_norm": 0.33266201615333557,
"learning_rate": 4.674581361782621e-05,
"loss": 1.3954,
"step": 6850
},
{
"epoch": 0.33,
"grad_norm": 0.4242490828037262,
"learning_rate": 4.673650836211634e-05,
"loss": 1.3416,
"step": 6860
},
{
"epoch": 0.33,
"grad_norm": 0.37850889563560486,
"learning_rate": 4.672719075022732e-05,
"loss": 1.2892,
"step": 6870
},
{
"epoch": 0.33,
"grad_norm": 0.4619060158729553,
"learning_rate": 4.671786078745578e-05,
"loss": 1.3431,
"step": 6880
},
{
"epoch": 0.33,
"grad_norm": 0.3185032606124878,
"learning_rate": 4.6708518479105346e-05,
"loss": 1.3351,
"step": 6890
},
{
"epoch": 0.33,
"grad_norm": 0.3289163112640381,
"learning_rate": 4.66991638304867e-05,
"loss": 1.3505,
"step": 6900
},
{
"epoch": 0.33,
"grad_norm": 0.4891585111618042,
"learning_rate": 4.66897968469175e-05,
"loss": 1.3663,
"step": 6910
},
{
"epoch": 0.33,
"grad_norm": 0.3460129201412201,
"learning_rate": 4.668041753372245e-05,
"loss": 1.3562,
"step": 6920
},
{
"epoch": 0.33,
"grad_norm": 0.3568609952926636,
"learning_rate": 4.6671965614423294e-05,
"loss": 1.2554,
"step": 6930
},
{
"epoch": 0.33,
"grad_norm": 0.5183925032615662,
"learning_rate": 4.666256288963373e-05,
"loss": 1.367,
"step": 6940
},
{
"epoch": 0.33,
"grad_norm": 0.33295902609825134,
"learning_rate": 4.665314785069952e-05,
"loss": 1.3217,
"step": 6950
},
{
"epoch": 0.33,
"grad_norm": 0.3377842605113983,
"learning_rate": 4.6643720502972674e-05,
"loss": 1.4117,
"step": 6960
},
{
"epoch": 0.33,
"grad_norm": 0.3380374014377594,
"learning_rate": 4.6634280851812206e-05,
"loss": 1.4206,
"step": 6970
},
{
"epoch": 0.34,
"grad_norm": 0.3161136209964752,
"learning_rate": 4.66248289025841e-05,
"loss": 1.3835,
"step": 6980
},
{
"epoch": 0.34,
"grad_norm": 0.30953165888786316,
"learning_rate": 4.661536466066135e-05,
"loss": 1.4828,
"step": 6990
},
{
"epoch": 0.34,
"grad_norm": 0.36132505536079407,
"learning_rate": 4.6605888131423925e-05,
"loss": 1.3959,
"step": 7000
},
{
"epoch": 0.34,
"grad_norm": 0.46117374300956726,
"learning_rate": 4.6596399320258784e-05,
"loss": 1.4304,
"step": 7010
},
{
"epoch": 0.34,
"grad_norm": 0.3435225188732147,
"learning_rate": 4.658689823255987e-05,
"loss": 1.3657,
"step": 7020
},
{
"epoch": 0.34,
"grad_norm": 0.45475250482559204,
"learning_rate": 4.6577384873728114e-05,
"loss": 1.4304,
"step": 7030
},
{
"epoch": 0.34,
"grad_norm": 0.3714883327484131,
"learning_rate": 4.656785924917139e-05,
"loss": 1.3655,
"step": 7040
},
{
"epoch": 0.34,
"grad_norm": 0.31743016839027405,
"learning_rate": 4.655832136430458e-05,
"loss": 1.4546,
"step": 7050
},
{
"epoch": 0.34,
"grad_norm": 0.3455131947994232,
"learning_rate": 4.654877122454952e-05,
"loss": 1.4078,
"step": 7060
},
{
"epoch": 0.34,
"grad_norm": 0.34969618916511536,
"learning_rate": 4.653920883533502e-05,
"loss": 1.3609,
"step": 7070
},
{
"epoch": 0.34,
"grad_norm": 0.4067552089691162,
"learning_rate": 4.652963420209684e-05,
"loss": 1.4395,
"step": 7080
},
{
"epoch": 0.34,
"grad_norm": 0.3989846706390381,
"learning_rate": 4.65200473302777e-05,
"loss": 1.3441,
"step": 7090
},
{
"epoch": 0.34,
"grad_norm": 0.4613402187824249,
"learning_rate": 4.651044822532729e-05,
"loss": 1.3504,
"step": 7100
},
{
"epoch": 0.34,
"grad_norm": 0.38455602526664734,
"learning_rate": 4.6500836892702245e-05,
"loss": 1.4855,
"step": 7110
},
{
"epoch": 0.34,
"grad_norm": 0.36579445004463196,
"learning_rate": 4.649121333786617e-05,
"loss": 1.2751,
"step": 7120
},
{
"epoch": 0.34,
"grad_norm": 0.3788816034793854,
"learning_rate": 4.648157756628959e-05,
"loss": 1.3796,
"step": 7130
},
{
"epoch": 0.34,
"grad_norm": 0.3098675608634949,
"learning_rate": 4.647192958344998e-05,
"loss": 1.3768,
"step": 7140
},
{
"epoch": 0.34,
"grad_norm": 0.3026945888996124,
"learning_rate": 4.646226939483177e-05,
"loss": 1.4164,
"step": 7150
},
{
"epoch": 0.34,
"grad_norm": 0.3602765202522278,
"learning_rate": 4.6452597005926325e-05,
"loss": 1.3392,
"step": 7160
},
{
"epoch": 0.34,
"grad_norm": 0.3817879557609558,
"learning_rate": 4.644291242223193e-05,
"loss": 1.4331,
"step": 7170
},
{
"epoch": 0.34,
"grad_norm": 0.3329647481441498,
"learning_rate": 4.6433215649253826e-05,
"loss": 1.4116,
"step": 7180
},
{
"epoch": 0.35,
"grad_norm": 0.4126259982585907,
"learning_rate": 4.642350669250417e-05,
"loss": 1.3795,
"step": 7190
},
{
"epoch": 0.35,
"grad_norm": 0.34714388847351074,
"learning_rate": 4.6413785557502034e-05,
"loss": 1.4108,
"step": 7200
},
{
"epoch": 0.35,
"grad_norm": 0.35611286759376526,
"learning_rate": 4.6404052249773435e-05,
"loss": 1.3009,
"step": 7210
},
{
"epoch": 0.35,
"grad_norm": 0.38419198989868164,
"learning_rate": 4.6394306774851294e-05,
"loss": 1.395,
"step": 7220
},
{
"epoch": 0.35,
"grad_norm": 0.36701419949531555,
"learning_rate": 4.6384549138275456e-05,
"loss": 1.3527,
"step": 7230
},
{
"epoch": 0.35,
"grad_norm": 0.3636476993560791,
"learning_rate": 4.637477934559268e-05,
"loss": 1.2547,
"step": 7240
},
{
"epoch": 0.35,
"grad_norm": 0.40688297152519226,
"learning_rate": 4.636499740235663e-05,
"loss": 1.3243,
"step": 7250
},
{
"epoch": 0.35,
"grad_norm": 0.5002498626708984,
"learning_rate": 4.635520331412787e-05,
"loss": 1.3278,
"step": 7260
},
{
"epoch": 0.35,
"grad_norm": 0.37266847491264343,
"learning_rate": 4.634539708647388e-05,
"loss": 1.5136,
"step": 7270
},
{
"epoch": 0.35,
"grad_norm": 0.3925212323665619,
"learning_rate": 4.633557872496905e-05,
"loss": 1.3568,
"step": 7280
},
{
"epoch": 0.35,
"grad_norm": 0.35610827803611755,
"learning_rate": 4.632574823519465e-05,
"loss": 1.3981,
"step": 7290
},
{
"epoch": 0.35,
"grad_norm": 0.35336118936538696,
"learning_rate": 4.631590562273884e-05,
"loss": 1.3158,
"step": 7300
},
{
"epoch": 0.35,
"grad_norm": 0.32741284370422363,
"learning_rate": 4.6306050893196694e-05,
"loss": 1.4064,
"step": 7310
},
{
"epoch": 0.35,
"grad_norm": 0.3675854206085205,
"learning_rate": 4.629618405217015e-05,
"loss": 1.4073,
"step": 7320
},
{
"epoch": 0.35,
"grad_norm": 0.35223251581192017,
"learning_rate": 4.6286305105268046e-05,
"loss": 1.3051,
"step": 7330
},
{
"epoch": 0.35,
"grad_norm": 0.375575453042984,
"learning_rate": 4.62764140581061e-05,
"loss": 1.3579,
"step": 7340
},
{
"epoch": 0.35,
"grad_norm": 0.5478109121322632,
"learning_rate": 4.6266510916306895e-05,
"loss": 1.2866,
"step": 7350
},
{
"epoch": 0.35,
"grad_norm": 0.32724764943122864,
"learning_rate": 4.6256595685499915e-05,
"loss": 1.3728,
"step": 7360
},
{
"epoch": 0.35,
"grad_norm": 0.4607096314430237,
"learning_rate": 4.624666837132149e-05,
"loss": 1.3714,
"step": 7370
},
{
"epoch": 0.35,
"grad_norm": 0.34265148639678955,
"learning_rate": 4.6236728979414826e-05,
"loss": 1.3445,
"step": 7380
},
{
"epoch": 0.35,
"grad_norm": 0.3650684356689453,
"learning_rate": 4.622677751543002e-05,
"loss": 1.3426,
"step": 7390
},
{
"epoch": 0.36,
"grad_norm": 0.33800172805786133,
"learning_rate": 4.6216813985023976e-05,
"loss": 1.3637,
"step": 7400
},
{
"epoch": 0.36,
"grad_norm": 0.6038261651992798,
"learning_rate": 4.620683839386052e-05,
"loss": 1.4593,
"step": 7410
},
{
"epoch": 0.36,
"grad_norm": 0.3368242681026459,
"learning_rate": 4.619685074761029e-05,
"loss": 1.3292,
"step": 7420
},
{
"epoch": 0.36,
"grad_norm": 0.38440167903900146,
"learning_rate": 4.618685105195079e-05,
"loss": 1.4249,
"step": 7430
},
{
"epoch": 0.36,
"grad_norm": 0.4154036343097687,
"learning_rate": 4.617683931256639e-05,
"loss": 1.3935,
"step": 7440
},
{
"epoch": 0.36,
"grad_norm": 0.3685297966003418,
"learning_rate": 4.6166815535148275e-05,
"loss": 1.3985,
"step": 7450
},
{
"epoch": 0.36,
"grad_norm": 0.396396666765213,
"learning_rate": 4.61567797253945e-05,
"loss": 1.3791,
"step": 7460
},
{
"epoch": 0.36,
"grad_norm": 0.45973828434944153,
"learning_rate": 4.6146731889009936e-05,
"loss": 1.3729,
"step": 7470
},
{
"epoch": 0.36,
"grad_norm": 0.3757419288158417,
"learning_rate": 4.613667203170632e-05,
"loss": 1.3888,
"step": 7480
},
{
"epoch": 0.36,
"grad_norm": 0.31412747502326965,
"learning_rate": 4.61266001592022e-05,
"loss": 1.3612,
"step": 7490
},
{
"epoch": 0.36,
"grad_norm": 0.36055684089660645,
"learning_rate": 4.6116516277222955e-05,
"loss": 1.3187,
"step": 7500
},
{
"epoch": 0.36,
"grad_norm": 0.34994739294052124,
"learning_rate": 4.61064203915008e-05,
"loss": 1.3706,
"step": 7510
},
{
"epoch": 0.36,
"grad_norm": 0.4439637362957001,
"learning_rate": 4.609631250777477e-05,
"loss": 1.4485,
"step": 7520
},
{
"epoch": 0.36,
"grad_norm": 0.4067482352256775,
"learning_rate": 4.6086192631790726e-05,
"loss": 1.313,
"step": 7530
},
{
"epoch": 0.36,
"grad_norm": 0.35776856541633606,
"learning_rate": 4.607606076930133e-05,
"loss": 1.3791,
"step": 7540
},
{
"epoch": 0.36,
"grad_norm": 0.4175155460834503,
"learning_rate": 4.6065916926066074e-05,
"loss": 1.438,
"step": 7550
},
{
"epoch": 0.36,
"grad_norm": 0.3096177279949188,
"learning_rate": 4.605576110785125e-05,
"loss": 1.4373,
"step": 7560
},
{
"epoch": 0.36,
"grad_norm": 0.4544389843940735,
"learning_rate": 4.604559332042996e-05,
"loss": 1.359,
"step": 7570
},
{
"epoch": 0.36,
"grad_norm": 0.3046031594276428,
"learning_rate": 4.6035413569582107e-05,
"loss": 1.3751,
"step": 7580
},
{
"epoch": 0.36,
"grad_norm": 0.35860735177993774,
"learning_rate": 4.602522186109442e-05,
"loss": 1.3349,
"step": 7590
},
{
"epoch": 0.36,
"grad_norm": 0.404680460691452,
"learning_rate": 4.601501820076038e-05,
"loss": 1.4903,
"step": 7600
},
{
"epoch": 0.37,
"grad_norm": 0.3761906325817108,
"learning_rate": 4.60048025943803e-05,
"loss": 1.4625,
"step": 7610
},
{
"epoch": 0.37,
"grad_norm": 0.3622777760028839,
"learning_rate": 4.599457504776127e-05,
"loss": 1.3991,
"step": 7620
},
{
"epoch": 0.37,
"grad_norm": 0.3617331385612488,
"learning_rate": 4.598433556671716e-05,
"loss": 1.3955,
"step": 7630
},
{
"epoch": 0.37,
"grad_norm": 0.4839256703853607,
"learning_rate": 4.5974084157068633e-05,
"loss": 1.3488,
"step": 7640
},
{
"epoch": 0.37,
"grad_norm": 0.40121495723724365,
"learning_rate": 4.596382082464313e-05,
"loss": 1.4109,
"step": 7650
},
{
"epoch": 0.37,
"grad_norm": 0.3255302608013153,
"learning_rate": 4.595354557527489e-05,
"loss": 1.302,
"step": 7660
},
{
"epoch": 0.37,
"grad_norm": 0.3410067558288574,
"learning_rate": 4.5943258414804885e-05,
"loss": 1.4285,
"step": 7670
},
{
"epoch": 0.37,
"grad_norm": 0.5079324245452881,
"learning_rate": 4.593295934908089e-05,
"loss": 1.3447,
"step": 7680
},
{
"epoch": 0.37,
"grad_norm": 0.34118178486824036,
"learning_rate": 4.592264838395744e-05,
"loss": 1.4016,
"step": 7690
},
{
"epoch": 0.37,
"grad_norm": 0.42170631885528564,
"learning_rate": 4.5912325525295816e-05,
"loss": 1.3888,
"step": 7700
},
{
"epoch": 0.37,
"grad_norm": 0.4335272014141083,
"learning_rate": 4.5901990778964093e-05,
"loss": 1.4144,
"step": 7710
},
{
"epoch": 0.37,
"grad_norm": 0.47610071301460266,
"learning_rate": 4.5891644150837086e-05,
"loss": 1.3226,
"step": 7720
},
{
"epoch": 0.37,
"grad_norm": 0.40044674277305603,
"learning_rate": 4.5881285646796354e-05,
"loss": 1.3907,
"step": 7730
},
{
"epoch": 0.37,
"grad_norm": 0.3251380920410156,
"learning_rate": 4.587091527273023e-05,
"loss": 1.4902,
"step": 7740
},
{
"epoch": 0.37,
"grad_norm": 0.5037009119987488,
"learning_rate": 4.586053303453377e-05,
"loss": 1.3129,
"step": 7750
},
{
"epoch": 0.37,
"grad_norm": 0.32583579421043396,
"learning_rate": 4.58501389381088e-05,
"loss": 1.418,
"step": 7760
},
{
"epoch": 0.37,
"grad_norm": 0.4100739359855652,
"learning_rate": 4.583973298936386e-05,
"loss": 1.3962,
"step": 7770
},
{
"epoch": 0.37,
"grad_norm": 0.34828880429267883,
"learning_rate": 4.582931519421424e-05,
"loss": 1.3986,
"step": 7780
},
{
"epoch": 0.37,
"grad_norm": 0.333126425743103,
"learning_rate": 4.5818885558581984e-05,
"loss": 1.3668,
"step": 7790
},
{
"epoch": 0.37,
"grad_norm": 0.3882949650287628,
"learning_rate": 4.580844408839583e-05,
"loss": 1.4071,
"step": 7800
},
{
"epoch": 0.37,
"grad_norm": 0.30676984786987305,
"learning_rate": 4.579799078959126e-05,
"loss": 1.3492,
"step": 7810
},
{
"epoch": 0.38,
"grad_norm": 0.47199365496635437,
"learning_rate": 4.5787525668110495e-05,
"loss": 1.3732,
"step": 7820
},
{
"epoch": 0.38,
"grad_norm": 0.35612753033638,
"learning_rate": 4.577704872990245e-05,
"loss": 1.3112,
"step": 7830
},
{
"epoch": 0.38,
"grad_norm": 0.40175312757492065,
"learning_rate": 4.5766559980922786e-05,
"loss": 1.401,
"step": 7840
},
{
"epoch": 0.38,
"grad_norm": 0.37230053544044495,
"learning_rate": 4.5756059427133835e-05,
"loss": 1.422,
"step": 7850
},
{
"epoch": 0.38,
"grad_norm": 0.34854793548583984,
"learning_rate": 4.574554707450469e-05,
"loss": 1.2693,
"step": 7860
},
{
"epoch": 0.38,
"grad_norm": 0.34508997201919556,
"learning_rate": 4.5735022929011115e-05,
"loss": 1.3235,
"step": 7870
},
{
"epoch": 0.38,
"grad_norm": 0.4071786105632782,
"learning_rate": 4.5724486996635595e-05,
"loss": 1.3977,
"step": 7880
},
{
"epoch": 0.38,
"grad_norm": 0.37492120265960693,
"learning_rate": 4.571393928336731e-05,
"loss": 1.3346,
"step": 7890
},
{
"epoch": 0.38,
"grad_norm": 0.3910008668899536,
"learning_rate": 4.570337979520214e-05,
"loss": 1.4166,
"step": 7900
},
{
"epoch": 0.38,
"grad_norm": 0.2957669794559479,
"learning_rate": 4.569280853814265e-05,
"loss": 1.3873,
"step": 7910
},
{
"epoch": 0.38,
"grad_norm": 0.4126977324485779,
"learning_rate": 4.568222551819811e-05,
"loss": 1.4232,
"step": 7920
},
{
"epoch": 0.38,
"grad_norm": 0.4155810475349426,
"learning_rate": 4.567163074138444e-05,
"loss": 1.3805,
"step": 7930
},
{
"epoch": 0.38,
"grad_norm": 0.3528597056865692,
"learning_rate": 4.566102421372431e-05,
"loss": 1.3275,
"step": 7940
},
{
"epoch": 0.38,
"grad_norm": 0.4245721399784088,
"learning_rate": 4.5650405941247005e-05,
"loss": 1.3206,
"step": 7950
},
{
"epoch": 0.38,
"grad_norm": 0.4109342694282532,
"learning_rate": 4.5639775929988516e-05,
"loss": 1.4058,
"step": 7960
},
{
"epoch": 0.38,
"grad_norm": 0.36114293336868286,
"learning_rate": 4.562913418599151e-05,
"loss": 1.3352,
"step": 7970
},
{
"epoch": 0.38,
"grad_norm": 0.44762012362480164,
"learning_rate": 4.5618480715305306e-05,
"loss": 1.3565,
"step": 7980
},
{
"epoch": 0.38,
"grad_norm": 0.3807506859302521,
"learning_rate": 4.5607815523985906e-05,
"loss": 1.3997,
"step": 7990
},
{
"epoch": 0.38,
"grad_norm": 0.3895118832588196,
"learning_rate": 4.559713861809597e-05,
"loss": 1.3767,
"step": 8000
},
{
"epoch": 0.38,
"grad_norm": 0.43535566329956055,
"learning_rate": 4.5586450003704805e-05,
"loss": 1.3664,
"step": 8010
},
{
"epoch": 0.38,
"grad_norm": 0.32971835136413574,
"learning_rate": 4.557574968688839e-05,
"loss": 1.2366,
"step": 8020
},
{
"epoch": 0.39,
"grad_norm": 0.42471224069595337,
"learning_rate": 4.556503767372935e-05,
"loss": 1.3709,
"step": 8030
},
{
"epoch": 0.39,
"grad_norm": 0.4666658043861389,
"learning_rate": 4.555431397031697e-05,
"loss": 1.3226,
"step": 8040
},
{
"epoch": 0.39,
"grad_norm": 0.31077998876571655,
"learning_rate": 4.554357858274715e-05,
"loss": 1.3323,
"step": 8050
},
{
"epoch": 0.39,
"grad_norm": 0.3165757358074188,
"learning_rate": 4.553283151712246e-05,
"loss": 1.3381,
"step": 8060
},
{
"epoch": 0.39,
"grad_norm": 0.3335585296154022,
"learning_rate": 4.55220727795521e-05,
"loss": 1.4317,
"step": 8070
},
{
"epoch": 0.39,
"grad_norm": 0.3384941816329956,
"learning_rate": 4.55113023761519e-05,
"loss": 1.3573,
"step": 8080
},
{
"epoch": 0.39,
"grad_norm": 0.3182198703289032,
"learning_rate": 4.550052031304434e-05,
"loss": 1.3059,
"step": 8090
},
{
"epoch": 0.39,
"grad_norm": 0.3508113622665405,
"learning_rate": 4.54897265963585e-05,
"loss": 1.3697,
"step": 8100
},
{
"epoch": 0.39,
"grad_norm": 0.3414270281791687,
"learning_rate": 4.5478921232230095e-05,
"loss": 1.352,
"step": 8110
},
{
"epoch": 0.39,
"grad_norm": 0.45592400431632996,
"learning_rate": 4.546810422680148e-05,
"loss": 1.4353,
"step": 8120
},
{
"epoch": 0.39,
"grad_norm": 0.32166656851768494,
"learning_rate": 4.545727558622159e-05,
"loss": 1.3925,
"step": 8130
},
{
"epoch": 0.39,
"grad_norm": 0.4481145441532135,
"learning_rate": 4.544643531664602e-05,
"loss": 1.3149,
"step": 8140
},
{
"epoch": 0.39,
"grad_norm": 0.3413998782634735,
"learning_rate": 4.543558342423695e-05,
"loss": 1.3638,
"step": 8150
},
{
"epoch": 0.39,
"grad_norm": 0.3290908634662628,
"learning_rate": 4.542471991516315e-05,
"loss": 1.2872,
"step": 8160
},
{
"epoch": 0.39,
"grad_norm": 0.3983047604560852,
"learning_rate": 4.5413844795600025e-05,
"loss": 1.4308,
"step": 8170
},
{
"epoch": 0.39,
"grad_norm": 0.3764418661594391,
"learning_rate": 4.5402958071729566e-05,
"loss": 1.3275,
"step": 8180
},
{
"epoch": 0.39,
"grad_norm": 0.5569051504135132,
"learning_rate": 4.539205974974036e-05,
"loss": 1.3825,
"step": 8190
},
{
"epoch": 0.39,
"grad_norm": 0.36069929599761963,
"learning_rate": 4.538114983582759e-05,
"loss": 1.3322,
"step": 8200
},
{
"epoch": 0.39,
"grad_norm": 0.32965072989463806,
"learning_rate": 4.5370228336193026e-05,
"loss": 1.3363,
"step": 8210
},
{
"epoch": 0.39,
"grad_norm": 0.3246593475341797,
"learning_rate": 4.535929525704503e-05,
"loss": 1.3792,
"step": 8220
},
{
"epoch": 0.4,
"grad_norm": 0.37570157647132874,
"learning_rate": 4.534835060459853e-05,
"loss": 1.3046,
"step": 8230
},
{
"epoch": 0.4,
"grad_norm": 0.3494696617126465,
"learning_rate": 4.533739438507505e-05,
"loss": 1.3998,
"step": 8240
},
{
"epoch": 0.4,
"grad_norm": 0.3457249701023102,
"learning_rate": 4.532642660470269e-05,
"loss": 1.3871,
"step": 8250
},
{
"epoch": 0.4,
"grad_norm": 0.5541589856147766,
"learning_rate": 4.531544726971612e-05,
"loss": 1.3754,
"step": 8260
},
{
"epoch": 0.4,
"grad_norm": 0.3757719397544861,
"learning_rate": 4.530445638635654e-05,
"loss": 1.3752,
"step": 8270
},
{
"epoch": 0.4,
"grad_norm": 0.30855974555015564,
"learning_rate": 4.529345396087179e-05,
"loss": 1.3404,
"step": 8280
},
{
"epoch": 0.4,
"grad_norm": 0.33023202419281006,
"learning_rate": 4.528243999951621e-05,
"loss": 1.2673,
"step": 8290
},
{
"epoch": 0.4,
"grad_norm": 0.361427366733551,
"learning_rate": 4.527141450855073e-05,
"loss": 1.3736,
"step": 8300
},
{
"epoch": 0.4,
"grad_norm": 0.32461586594581604,
"learning_rate": 4.52603774942428e-05,
"loss": 1.2827,
"step": 8310
},
{
"epoch": 0.4,
"grad_norm": 0.2874002754688263,
"learning_rate": 4.524932896286644e-05,
"loss": 1.4151,
"step": 8320
},
{
"epoch": 0.4,
"grad_norm": 0.39023908972740173,
"learning_rate": 4.523826892070225e-05,
"loss": 1.4004,
"step": 8330
},
{
"epoch": 0.4,
"grad_norm": 0.38757333159446716,
"learning_rate": 4.522719737403732e-05,
"loss": 1.3533,
"step": 8340
},
{
"epoch": 0.4,
"grad_norm": 0.3636305034160614,
"learning_rate": 4.5216114329165305e-05,
"loss": 1.2994,
"step": 8350
},
{
"epoch": 0.4,
"grad_norm": 0.42119061946868896,
"learning_rate": 4.52050197923864e-05,
"loss": 1.3234,
"step": 8360
},
{
"epoch": 0.4,
"grad_norm": 0.42275360226631165,
"learning_rate": 4.51939137700073e-05,
"loss": 1.3349,
"step": 8370
},
{
"epoch": 0.4,
"grad_norm": 0.3990207314491272,
"learning_rate": 4.518279626834129e-05,
"loss": 1.3401,
"step": 8380
},
{
"epoch": 0.4,
"grad_norm": 0.4044749140739441,
"learning_rate": 4.517166729370812e-05,
"loss": 1.3729,
"step": 8390
},
{
"epoch": 0.4,
"grad_norm": 0.44323498010635376,
"learning_rate": 4.516052685243409e-05,
"loss": 1.362,
"step": 8400
},
{
"epoch": 0.4,
"grad_norm": 0.3554174602031708,
"learning_rate": 4.514937495085203e-05,
"loss": 1.3703,
"step": 8410
},
{
"epoch": 0.4,
"grad_norm": 0.4610666036605835,
"learning_rate": 4.513821159530126e-05,
"loss": 1.4329,
"step": 8420
},
{
"epoch": 0.4,
"grad_norm": 0.3678601086139679,
"learning_rate": 4.5127036792127616e-05,
"loss": 1.3806,
"step": 8430
},
{
"epoch": 0.41,
"grad_norm": 0.41576239466667175,
"learning_rate": 4.5115850547683456e-05,
"loss": 1.4026,
"step": 8440
},
{
"epoch": 0.41,
"grad_norm": 0.3420347571372986,
"learning_rate": 4.5104652868327624e-05,
"loss": 1.4146,
"step": 8450
},
{
"epoch": 0.41,
"grad_norm": 0.4109761416912079,
"learning_rate": 4.509344376042548e-05,
"loss": 1.3876,
"step": 8460
},
{
"epoch": 0.41,
"grad_norm": 0.43770748376846313,
"learning_rate": 4.508222323034886e-05,
"loss": 1.4394,
"step": 8470
},
{
"epoch": 0.41,
"grad_norm": 0.4325715899467468,
"learning_rate": 4.5070991284476115e-05,
"loss": 1.3058,
"step": 8480
},
{
"epoch": 0.41,
"grad_norm": 0.40183165669441223,
"learning_rate": 4.5059747929192074e-05,
"loss": 1.3289,
"step": 8490
},
{
"epoch": 0.41,
"grad_norm": 0.43909403681755066,
"learning_rate": 4.5048493170888065e-05,
"loss": 1.2683,
"step": 8500
},
{
"epoch": 0.41,
"grad_norm": 0.36718305945396423,
"learning_rate": 4.503722701596187e-05,
"loss": 1.395,
"step": 8510
},
{
"epoch": 0.41,
"grad_norm": 0.41779977083206177,
"learning_rate": 4.502594947081777e-05,
"loss": 1.3765,
"step": 8520
},
{
"epoch": 0.41,
"grad_norm": 0.422727108001709,
"learning_rate": 4.5014660541866524e-05,
"loss": 1.3532,
"step": 8530
},
{
"epoch": 0.41,
"grad_norm": 0.4417511224746704,
"learning_rate": 4.500336023552535e-05,
"loss": 1.3453,
"step": 8540
},
{
"epoch": 0.41,
"grad_norm": 0.43296435475349426,
"learning_rate": 4.4992048558217945e-05,
"loss": 1.4395,
"step": 8550
},
{
"epoch": 0.41,
"grad_norm": 0.5378210544586182,
"learning_rate": 4.4980725516374456e-05,
"loss": 1.3385,
"step": 8560
},
{
"epoch": 0.41,
"grad_norm": 0.37314531207084656,
"learning_rate": 4.4969391116431494e-05,
"loss": 1.2799,
"step": 8570
},
{
"epoch": 0.41,
"grad_norm": 0.3360219895839691,
"learning_rate": 4.495804536483215e-05,
"loss": 1.3951,
"step": 8580
},
{
"epoch": 0.41,
"grad_norm": 0.33808451890945435,
"learning_rate": 4.4946688268025926e-05,
"loss": 1.3161,
"step": 8590
},
{
"epoch": 0.41,
"grad_norm": 0.4306478500366211,
"learning_rate": 4.493531983246881e-05,
"loss": 1.3401,
"step": 8600
},
{
"epoch": 0.41,
"grad_norm": 0.32607635855674744,
"learning_rate": 4.4923940064623215e-05,
"loss": 1.3583,
"step": 8610
},
{
"epoch": 0.41,
"grad_norm": 0.3098576068878174,
"learning_rate": 4.4912548970958e-05,
"loss": 1.3485,
"step": 8620
},
{
"epoch": 0.41,
"grad_norm": 0.3063451647758484,
"learning_rate": 4.490114655794848e-05,
"loss": 1.3833,
"step": 8630
},
{
"epoch": 0.41,
"grad_norm": 0.3353252708911896,
"learning_rate": 4.4889732832076364e-05,
"loss": 1.354,
"step": 8640
},
{
"epoch": 0.42,
"grad_norm": 0.38628441095352173,
"learning_rate": 4.487830779982983e-05,
"loss": 1.4336,
"step": 8650
},
{
"epoch": 0.42,
"grad_norm": 0.4486022889614105,
"learning_rate": 4.4866871467703475e-05,
"loss": 1.4349,
"step": 8660
},
{
"epoch": 0.42,
"grad_norm": 0.4411972463130951,
"learning_rate": 4.48554238421983e-05,
"loss": 1.4359,
"step": 8670
},
{
"epoch": 0.42,
"grad_norm": 0.4130594730377197,
"learning_rate": 4.484396492982175e-05,
"loss": 1.3774,
"step": 8680
},
{
"epoch": 0.42,
"grad_norm": 0.463099867105484,
"learning_rate": 4.4832494737087676e-05,
"loss": 1.3583,
"step": 8690
},
{
"epoch": 0.42,
"grad_norm": 0.3544412851333618,
"learning_rate": 4.4821013270516335e-05,
"loss": 1.3255,
"step": 8700
},
{
"epoch": 0.42,
"grad_norm": 0.38806211948394775,
"learning_rate": 4.480952053663441e-05,
"loss": 1.3516,
"step": 8710
},
{
"epoch": 0.42,
"grad_norm": 0.3644128739833832,
"learning_rate": 4.479801654197496e-05,
"loss": 1.3176,
"step": 8720
},
{
"epoch": 0.42,
"grad_norm": 0.3879926800727844,
"learning_rate": 4.478650129307748e-05,
"loss": 1.3562,
"step": 8730
},
{
"epoch": 0.42,
"grad_norm": 0.34438619017601013,
"learning_rate": 4.477497479648783e-05,
"loss": 1.2998,
"step": 8740
},
{
"epoch": 0.42,
"grad_norm": 0.3750459849834442,
"learning_rate": 4.476343705875829e-05,
"loss": 1.371,
"step": 8750
},
{
"epoch": 0.42,
"grad_norm": 0.37031787633895874,
"learning_rate": 4.475188808644753e-05,
"loss": 1.3773,
"step": 8760
},
{
"epoch": 0.42,
"grad_norm": 0.4215640127658844,
"learning_rate": 4.4740327886120566e-05,
"loss": 1.395,
"step": 8770
},
{
"epoch": 0.42,
"grad_norm": 0.33745723962783813,
"learning_rate": 4.4728756464348846e-05,
"loss": 1.428,
"step": 8780
},
{
"epoch": 0.42,
"grad_norm": 0.4800817370414734,
"learning_rate": 4.471717382771018e-05,
"loss": 1.3225,
"step": 8790
},
{
"epoch": 0.42,
"grad_norm": 0.47044068574905396,
"learning_rate": 4.470557998278875e-05,
"loss": 1.3861,
"step": 8800
},
{
"epoch": 0.42,
"grad_norm": 0.40047499537467957,
"learning_rate": 4.469397493617509e-05,
"loss": 1.446,
"step": 8810
},
{
"epoch": 0.42,
"grad_norm": 0.3730560541152954,
"learning_rate": 4.468235869446614e-05,
"loss": 1.3095,
"step": 8820
},
{
"epoch": 0.42,
"grad_norm": 0.3734876215457916,
"learning_rate": 4.467073126426519e-05,
"loss": 1.3765,
"step": 8830
},
{
"epoch": 0.42,
"grad_norm": 0.43869125843048096,
"learning_rate": 4.465909265218187e-05,
"loss": 1.3742,
"step": 8840
},
{
"epoch": 0.42,
"grad_norm": 0.3163073658943176,
"learning_rate": 4.464744286483219e-05,
"loss": 1.3291,
"step": 8850
},
{
"epoch": 0.43,
"grad_norm": 0.30170637369155884,
"learning_rate": 4.46357819088385e-05,
"loss": 1.3608,
"step": 8860
},
{
"epoch": 0.43,
"grad_norm": 0.3661230802536011,
"learning_rate": 4.46241097908295e-05,
"loss": 1.4034,
"step": 8870
},
{
"epoch": 0.43,
"grad_norm": 0.49004048109054565,
"learning_rate": 4.4612426517440255e-05,
"loss": 1.3856,
"step": 8880
},
{
"epoch": 0.43,
"grad_norm": 0.3834657669067383,
"learning_rate": 4.4600732095312134e-05,
"loss": 1.3708,
"step": 8890
},
{
"epoch": 0.43,
"grad_norm": 0.40286943316459656,
"learning_rate": 4.4589026531092884e-05,
"loss": 1.3805,
"step": 8900
},
{
"epoch": 0.43,
"grad_norm": 0.37186625599861145,
"learning_rate": 4.4577309831436545e-05,
"loss": 1.3779,
"step": 8910
},
{
"epoch": 0.43,
"grad_norm": 0.3426941931247711,
"learning_rate": 4.456558200300352e-05,
"loss": 1.2864,
"step": 8920
},
{
"epoch": 0.43,
"grad_norm": 0.494442343711853,
"learning_rate": 4.455501744781964e-05,
"loss": 1.422,
"step": 8930
},
{
"epoch": 0.43,
"grad_norm": 0.34884506464004517,
"learning_rate": 4.454326849308294e-05,
"loss": 1.3128,
"step": 8940
},
{
"epoch": 0.43,
"grad_norm": 0.35986900329589844,
"learning_rate": 4.453150842892044e-05,
"loss": 1.3486,
"step": 8950
},
{
"epoch": 0.43,
"grad_norm": 0.3149203658103943,
"learning_rate": 4.4519737262017195e-05,
"loss": 1.2959,
"step": 8960
},
{
"epoch": 0.43,
"grad_norm": 0.3722383677959442,
"learning_rate": 4.450795499906455e-05,
"loss": 1.4113,
"step": 8970
},
{
"epoch": 0.43,
"grad_norm": 0.291980504989624,
"learning_rate": 4.449734148082042e-05,
"loss": 1.3277,
"step": 8980
},
{
"epoch": 0.43,
"grad_norm": 0.5708626508712769,
"learning_rate": 4.4485538153831175e-05,
"loss": 1.4262,
"step": 8990
},
{
"epoch": 0.43,
"grad_norm": 0.33955106139183044,
"learning_rate": 4.447372375023311e-05,
"loss": 1.3401,
"step": 9000
},
{
"epoch": 0.43,
"grad_norm": 0.4389679431915283,
"learning_rate": 4.446189827674216e-05,
"loss": 1.4039,
"step": 9010
},
{
"epoch": 0.43,
"grad_norm": 0.41198277473449707,
"learning_rate": 4.445006174008054e-05,
"loss": 1.365,
"step": 9020
},
{
"epoch": 0.43,
"grad_norm": 0.37900421023368835,
"learning_rate": 4.4438214146976767e-05,
"loss": 1.377,
"step": 9030
},
{
"epoch": 0.43,
"grad_norm": 0.3433659076690674,
"learning_rate": 4.442635550416564e-05,
"loss": 1.3474,
"step": 9040
},
{
"epoch": 0.43,
"grad_norm": 0.4139823615550995,
"learning_rate": 4.441448581838823e-05,
"loss": 1.3431,
"step": 9050
},
{
"epoch": 0.43,
"grad_norm": 0.4547215402126312,
"learning_rate": 4.44026050963919e-05,
"loss": 1.3456,
"step": 9060
},
{
"epoch": 0.44,
"grad_norm": 0.37096837162971497,
"learning_rate": 4.4390713344930285e-05,
"loss": 1.4243,
"step": 9070
},
{
"epoch": 0.44,
"grad_norm": 0.343124657869339,
"learning_rate": 4.4378810570763275e-05,
"loss": 1.4023,
"step": 9080
},
{
"epoch": 0.44,
"grad_norm": 0.4370552599430084,
"learning_rate": 4.436689678065703e-05,
"loss": 1.3941,
"step": 9090
},
{
"epoch": 0.44,
"grad_norm": 0.4587309658527374,
"learning_rate": 4.4354971981384e-05,
"loss": 1.4327,
"step": 9100
},
{
"epoch": 0.44,
"grad_norm": 0.42083248496055603,
"learning_rate": 4.4343036179722845e-05,
"loss": 1.323,
"step": 9110
},
{
"epoch": 0.44,
"grad_norm": 0.3054110109806061,
"learning_rate": 4.4331089382458527e-05,
"loss": 1.4616,
"step": 9120
},
{
"epoch": 0.44,
"grad_norm": 0.3561241924762726,
"learning_rate": 4.4319131596382216e-05,
"loss": 1.4109,
"step": 9130
},
{
"epoch": 0.44,
"grad_norm": 0.38903114199638367,
"learning_rate": 4.430716282829135e-05,
"loss": 1.3716,
"step": 9140
},
{
"epoch": 0.44,
"grad_norm": 0.27901893854141235,
"learning_rate": 4.4295183084989635e-05,
"loss": 1.3656,
"step": 9150
},
{
"epoch": 0.44,
"grad_norm": 0.3753771185874939,
"learning_rate": 4.428319237328696e-05,
"loss": 1.3345,
"step": 9160
},
{
"epoch": 0.44,
"grad_norm": 0.63884437084198,
"learning_rate": 4.42711906999995e-05,
"loss": 1.3767,
"step": 9170
},
{
"epoch": 0.44,
"grad_norm": 0.3387126624584198,
"learning_rate": 4.4259178071949626e-05,
"loss": 1.3706,
"step": 9180
},
{
"epoch": 0.44,
"grad_norm": 0.37534603476524353,
"learning_rate": 4.424715449596596e-05,
"loss": 1.3555,
"step": 9190
},
{
"epoch": 0.44,
"grad_norm": 0.40485525131225586,
"learning_rate": 4.423511997888333e-05,
"loss": 1.2897,
"step": 9200
},
{
"epoch": 0.44,
"grad_norm": 0.33167576789855957,
"learning_rate": 4.422307452754279e-05,
"loss": 1.3912,
"step": 9210
},
{
"epoch": 0.44,
"grad_norm": 0.3173292875289917,
"learning_rate": 4.421101814879161e-05,
"loss": 1.3293,
"step": 9220
},
{
"epoch": 0.44,
"grad_norm": 0.4590155780315399,
"learning_rate": 4.419895084948329e-05,
"loss": 1.2925,
"step": 9230
},
{
"epoch": 0.44,
"grad_norm": 0.3406144082546234,
"learning_rate": 4.4186872636477496e-05,
"loss": 1.2975,
"step": 9240
},
{
"epoch": 0.44,
"grad_norm": 0.5175010561943054,
"learning_rate": 4.417478351664013e-05,
"loss": 1.339,
"step": 9250
},
{
"epoch": 0.44,
"grad_norm": 0.37617480754852295,
"learning_rate": 4.416268349684329e-05,
"loss": 1.3124,
"step": 9260
},
{
"epoch": 0.44,
"grad_norm": 0.38344213366508484,
"learning_rate": 4.415057258396528e-05,
"loss": 1.31,
"step": 9270
},
{
"epoch": 0.45,
"grad_norm": 0.3458814024925232,
"learning_rate": 4.413845078489055e-05,
"loss": 1.3999,
"step": 9280
},
{
"epoch": 0.45,
"grad_norm": 0.4557632505893707,
"learning_rate": 4.412631810650979e-05,
"loss": 1.3355,
"step": 9290
},
{
"epoch": 0.45,
"grad_norm": 0.3213663101196289,
"learning_rate": 4.411417455571986e-05,
"loss": 1.3043,
"step": 9300
},
{
"epoch": 0.45,
"grad_norm": 0.4033249318599701,
"learning_rate": 4.4102020139423785e-05,
"loss": 1.3575,
"step": 9310
},
{
"epoch": 0.45,
"grad_norm": 0.3848459720611572,
"learning_rate": 4.408985486453078e-05,
"loss": 1.3617,
"step": 9320
},
{
"epoch": 0.45,
"grad_norm": 0.32192572951316833,
"learning_rate": 4.4077678737956225e-05,
"loss": 1.4015,
"step": 9330
},
{
"epoch": 0.45,
"grad_norm": 0.4720775783061981,
"learning_rate": 4.406549176662168e-05,
"loss": 1.3025,
"step": 9340
},
{
"epoch": 0.45,
"grad_norm": 0.4061540961265564,
"learning_rate": 4.405329395745487e-05,
"loss": 1.2517,
"step": 9350
},
{
"epoch": 0.45,
"grad_norm": 0.34050253033638,
"learning_rate": 4.404108531738965e-05,
"loss": 1.4707,
"step": 9360
},
{
"epoch": 0.45,
"grad_norm": 0.38199788331985474,
"learning_rate": 4.402886585336608e-05,
"loss": 1.3035,
"step": 9370
},
{
"epoch": 0.45,
"grad_norm": 0.649121880531311,
"learning_rate": 4.401663557233035e-05,
"loss": 1.2984,
"step": 9380
},
{
"epoch": 0.45,
"grad_norm": 0.3654865026473999,
"learning_rate": 4.400439448123477e-05,
"loss": 1.3217,
"step": 9390
},
{
"epoch": 0.45,
"grad_norm": 0.4124996066093445,
"learning_rate": 4.399214258703784e-05,
"loss": 1.4144,
"step": 9400
},
{
"epoch": 0.45,
"grad_norm": 0.3885655403137207,
"learning_rate": 4.397987989670419e-05,
"loss": 1.3382,
"step": 9410
},
{
"epoch": 0.45,
"grad_norm": 0.40180638432502747,
"learning_rate": 4.396760641720458e-05,
"loss": 1.3484,
"step": 9420
},
{
"epoch": 0.45,
"grad_norm": 0.3806428015232086,
"learning_rate": 4.395532215551589e-05,
"loss": 1.3505,
"step": 9430
},
{
"epoch": 0.45,
"grad_norm": 0.4138786494731903,
"learning_rate": 4.3943027118621147e-05,
"loss": 1.376,
"step": 9440
},
{
"epoch": 0.45,
"grad_norm": 0.4543648362159729,
"learning_rate": 4.3930721313509506e-05,
"loss": 1.3921,
"step": 9450
},
{
"epoch": 0.45,
"grad_norm": 0.36886969208717346,
"learning_rate": 4.3918404747176234e-05,
"loss": 1.3678,
"step": 9460
},
{
"epoch": 0.45,
"grad_norm": 0.5785180330276489,
"learning_rate": 4.390607742662272e-05,
"loss": 1.2747,
"step": 9470
},
{
"epoch": 0.46,
"grad_norm": 0.43989765644073486,
"learning_rate": 4.389373935885646e-05,
"loss": 1.3355,
"step": 9480
},
{
"epoch": 0.46,
"grad_norm": 0.34230470657348633,
"learning_rate": 4.388139055089107e-05,
"loss": 1.4013,
"step": 9490
},
{
"epoch": 0.46,
"grad_norm": 0.4549820125102997,
"learning_rate": 4.386903100974624e-05,
"loss": 1.4941,
"step": 9500
},
{
"epoch": 0.46,
"grad_norm": 0.3718874156475067,
"learning_rate": 4.385666074244782e-05,
"loss": 1.3817,
"step": 9510
},
{
"epoch": 0.46,
"grad_norm": 0.5056178569793701,
"learning_rate": 4.3844279756027706e-05,
"loss": 1.4725,
"step": 9520
},
{
"epoch": 0.46,
"grad_norm": 0.5005752444267273,
"learning_rate": 4.383188805752391e-05,
"loss": 1.3153,
"step": 9530
},
{
"epoch": 0.46,
"grad_norm": 0.4389950931072235,
"learning_rate": 4.382072637586076e-05,
"loss": 1.4412,
"step": 9540
},
{
"epoch": 0.46,
"grad_norm": 0.3834897577762604,
"learning_rate": 4.3808314343809464e-05,
"loss": 1.3541,
"step": 9550
},
{
"epoch": 0.46,
"grad_norm": 0.38238221406936646,
"learning_rate": 4.379589162011912e-05,
"loss": 1.3597,
"step": 9560
},
{
"epoch": 0.46,
"grad_norm": 0.4164634048938751,
"learning_rate": 4.378345821185148e-05,
"loss": 1.3389,
"step": 9570
},
{
"epoch": 0.46,
"grad_norm": 0.4651401937007904,
"learning_rate": 4.377101412607432e-05,
"loss": 1.4178,
"step": 9580
},
{
"epoch": 0.46,
"grad_norm": 0.34688007831573486,
"learning_rate": 4.375855936986153e-05,
"loss": 1.4524,
"step": 9590
},
{
"epoch": 0.46,
"grad_norm": 0.4452615976333618,
"learning_rate": 4.3746093950293045e-05,
"loss": 1.3992,
"step": 9600
},
{
"epoch": 0.46,
"grad_norm": 0.4621642529964447,
"learning_rate": 4.373361787445487e-05,
"loss": 1.2887,
"step": 9610
},
{
"epoch": 0.46,
"grad_norm": 0.3862049877643585,
"learning_rate": 4.3721131149439064e-05,
"loss": 1.2666,
"step": 9620
},
{
"epoch": 0.46,
"grad_norm": 0.3947352170944214,
"learning_rate": 4.370863378234375e-05,
"loss": 1.3141,
"step": 9630
},
{
"epoch": 0.46,
"grad_norm": 0.32293689250946045,
"learning_rate": 4.369612578027308e-05,
"loss": 1.3321,
"step": 9640
},
{
"epoch": 0.46,
"grad_norm": 0.38175705075263977,
"learning_rate": 4.368360715033727e-05,
"loss": 1.318,
"step": 9650
},
{
"epoch": 0.46,
"grad_norm": 0.45248037576675415,
"learning_rate": 4.3671077899652565e-05,
"loss": 1.2806,
"step": 9660
},
{
"epoch": 0.46,
"grad_norm": 0.5387378334999084,
"learning_rate": 4.365853803534126e-05,
"loss": 1.3717,
"step": 9670
},
{
"epoch": 0.46,
"grad_norm": 0.29943278431892395,
"learning_rate": 4.364598756453167e-05,
"loss": 1.2897,
"step": 9680
},
{
"epoch": 0.47,
"grad_norm": 0.420290470123291,
"learning_rate": 4.363342649435814e-05,
"loss": 1.3149,
"step": 9690
},
{
"epoch": 0.47,
"grad_norm": 0.3173505663871765,
"learning_rate": 4.362085483196107e-05,
"loss": 1.2487,
"step": 9700
},
{
"epoch": 0.47,
"grad_norm": 0.45538169145584106,
"learning_rate": 4.360827258448682e-05,
"loss": 1.3602,
"step": 9710
},
{
"epoch": 0.47,
"grad_norm": 0.38612720370292664,
"learning_rate": 4.359567975908783e-05,
"loss": 1.3058,
"step": 9720
},
{
"epoch": 0.47,
"grad_norm": 0.33554235100746155,
"learning_rate": 4.358307636292252e-05,
"loss": 1.3744,
"step": 9730
},
{
"epoch": 0.47,
"grad_norm": 0.4160993993282318,
"learning_rate": 4.357046240315533e-05,
"loss": 1.3394,
"step": 9740
},
{
"epoch": 0.47,
"grad_norm": 0.43627703189849854,
"learning_rate": 4.3557837886956686e-05,
"loss": 1.3521,
"step": 9750
},
{
"epoch": 0.47,
"grad_norm": 0.5952196717262268,
"learning_rate": 4.354520282150305e-05,
"loss": 1.2138,
"step": 9760
},
{
"epoch": 0.47,
"grad_norm": 0.41586601734161377,
"learning_rate": 4.3532557213976846e-05,
"loss": 1.3528,
"step": 9770
},
{
"epoch": 0.47,
"grad_norm": 0.33763739466667175,
"learning_rate": 4.351990107156651e-05,
"loss": 1.32,
"step": 9780
},
{
"epoch": 0.47,
"grad_norm": 0.34863564372062683,
"learning_rate": 4.350723440146646e-05,
"loss": 1.325,
"step": 9790
},
{
"epoch": 0.47,
"grad_norm": 0.35480672121047974,
"learning_rate": 4.349455721087711e-05,
"loss": 1.4571,
"step": 9800
},
{
"epoch": 0.47,
"grad_norm": 0.33991697430610657,
"learning_rate": 4.348186950700482e-05,
"loss": 1.3646,
"step": 9810
},
{
"epoch": 0.47,
"grad_norm": 0.324405699968338,
"learning_rate": 4.3469171297061984e-05,
"loss": 1.3203,
"step": 9820
},
{
"epoch": 0.47,
"grad_norm": 0.36518266797065735,
"learning_rate": 4.345646258826691e-05,
"loss": 1.3238,
"step": 9830
},
{
"epoch": 0.47,
"grad_norm": 0.47166916728019714,
"learning_rate": 4.344374338784391e-05,
"loss": 1.3201,
"step": 9840
},
{
"epoch": 0.47,
"grad_norm": 0.3878999352455139,
"learning_rate": 4.343101370302324e-05,
"loss": 1.3158,
"step": 9850
},
{
"epoch": 0.47,
"grad_norm": 0.38008835911750793,
"learning_rate": 4.341827354104113e-05,
"loss": 1.3044,
"step": 9860
},
{
"epoch": 0.47,
"grad_norm": 0.4728231430053711,
"learning_rate": 4.340552290913977e-05,
"loss": 1.4093,
"step": 9870
},
{
"epoch": 0.47,
"grad_norm": 0.5555732250213623,
"learning_rate": 4.339276181456728e-05,
"loss": 1.2725,
"step": 9880
},
{
"epoch": 0.47,
"grad_norm": 0.4192439913749695,
"learning_rate": 4.3379990264577755e-05,
"loss": 1.3977,
"step": 9890
},
{
"epoch": 0.48,
"grad_norm": 0.4367178976535797,
"learning_rate": 4.3367208266431194e-05,
"loss": 1.3307,
"step": 9900
},
{
"epoch": 0.48,
"grad_norm": 0.33396115899086,
"learning_rate": 4.335441582739358e-05,
"loss": 1.3987,
"step": 9910
},
{
"epoch": 0.48,
"grad_norm": 0.5095841288566589,
"learning_rate": 4.3341612954736796e-05,
"loss": 1.2933,
"step": 9920
},
{
"epoch": 0.48,
"grad_norm": 0.3360755741596222,
"learning_rate": 4.3328799655738696e-05,
"loss": 1.3432,
"step": 9930
},
{
"epoch": 0.48,
"grad_norm": 0.40567177534103394,
"learning_rate": 4.3315975937683004e-05,
"loss": 1.3514,
"step": 9940
},
{
"epoch": 0.48,
"grad_norm": 0.42016854882240295,
"learning_rate": 4.330314180785943e-05,
"loss": 1.4049,
"step": 9950
},
{
"epoch": 0.48,
"grad_norm": 0.4038058817386627,
"learning_rate": 4.329029727356354e-05,
"loss": 1.4175,
"step": 9960
},
{
"epoch": 0.48,
"grad_norm": 0.5521274209022522,
"learning_rate": 4.327744234209686e-05,
"loss": 1.4322,
"step": 9970
},
{
"epoch": 0.48,
"grad_norm": 0.4091934859752655,
"learning_rate": 4.326457702076682e-05,
"loss": 1.3298,
"step": 9980
},
{
"epoch": 0.48,
"grad_norm": 0.3261078894138336,
"learning_rate": 4.325170131688674e-05,
"loss": 1.4278,
"step": 9990
},
{
"epoch": 0.48,
"grad_norm": 0.4421689808368683,
"learning_rate": 4.323881523777584e-05,
"loss": 1.3656,
"step": 10000
},
{
"epoch": 0.48,
"grad_norm": 0.3652206063270569,
"learning_rate": 4.3225918790759275e-05,
"loss": 1.3579,
"step": 10010
},
{
"epoch": 0.48,
"grad_norm": 0.34313803911209106,
"learning_rate": 4.321301198316804e-05,
"loss": 1.3734,
"step": 10020
},
{
"epoch": 0.48,
"grad_norm": 0.3687114417552948,
"learning_rate": 4.3200094822339054e-05,
"loss": 1.3613,
"step": 10030
},
{
"epoch": 0.48,
"grad_norm": 0.30381107330322266,
"learning_rate": 4.318716731561512e-05,
"loss": 1.363,
"step": 10040
},
{
"epoch": 0.48,
"grad_norm": 0.45174625515937805,
"learning_rate": 4.3174229470344904e-05,
"loss": 1.403,
"step": 10050
},
{
"epoch": 0.48,
"grad_norm": 0.46961259841918945,
"learning_rate": 4.316128129388297e-05,
"loss": 1.4585,
"step": 10060
},
{
"epoch": 0.48,
"grad_norm": 0.42780056595802307,
"learning_rate": 4.314832279358974e-05,
"loss": 1.2452,
"step": 10070
},
{
"epoch": 0.48,
"grad_norm": 0.3768773376941681,
"learning_rate": 4.3135353976831516e-05,
"loss": 1.4483,
"step": 10080
},
{
"epoch": 0.48,
"grad_norm": 0.44627833366394043,
"learning_rate": 4.3122374850980454e-05,
"loss": 1.3804,
"step": 10090
},
{
"epoch": 0.48,
"grad_norm": 0.3656626343727112,
"learning_rate": 4.310938542341456e-05,
"loss": 1.3529,
"step": 10100
},
{
"epoch": 0.49,
"grad_norm": 0.3566610515117645,
"learning_rate": 4.3096385701517744e-05,
"loss": 1.3487,
"step": 10110
},
{
"epoch": 0.49,
"grad_norm": 0.3319740295410156,
"learning_rate": 4.3083375692679705e-05,
"loss": 1.4854,
"step": 10120
},
{
"epoch": 0.49,
"grad_norm": 0.4777064025402069,
"learning_rate": 4.307035540429604e-05,
"loss": 1.3638,
"step": 10130
},
{
"epoch": 0.49,
"grad_norm": 0.3526528775691986,
"learning_rate": 4.305732484376814e-05,
"loss": 1.4037,
"step": 10140
},
{
"epoch": 0.49,
"grad_norm": 0.37839627265930176,
"learning_rate": 4.304428401850329e-05,
"loss": 1.3633,
"step": 10150
},
{
"epoch": 0.49,
"grad_norm": 0.40401017665863037,
"learning_rate": 4.3031232935914584e-05,
"loss": 1.3297,
"step": 10160
},
{
"epoch": 0.49,
"grad_norm": 0.47318682074546814,
"learning_rate": 4.3018171603420934e-05,
"loss": 1.3713,
"step": 10170
},
{
"epoch": 0.49,
"grad_norm": 0.32703691720962524,
"learning_rate": 4.3005100028447086e-05,
"loss": 1.3657,
"step": 10180
},
{
"epoch": 0.49,
"grad_norm": 0.5559362173080444,
"learning_rate": 4.299201821842363e-05,
"loss": 1.4275,
"step": 10190
},
{
"epoch": 0.49,
"grad_norm": 0.4326966106891632,
"learning_rate": 4.297892618078695e-05,
"loss": 1.3078,
"step": 10200
},
{
"epoch": 0.49,
"grad_norm": 0.3555945158004761,
"learning_rate": 4.296582392297925e-05,
"loss": 1.3538,
"step": 10210
},
{
"epoch": 0.49,
"grad_norm": 0.38792574405670166,
"learning_rate": 4.295271145244855e-05,
"loss": 1.3011,
"step": 10220
},
{
"epoch": 0.49,
"grad_norm": 0.285739928483963,
"learning_rate": 4.2939588776648655e-05,
"loss": 1.3792,
"step": 10230
},
{
"epoch": 0.49,
"grad_norm": 0.38338321447372437,
"learning_rate": 4.2926455903039206e-05,
"loss": 1.3961,
"step": 10240
},
{
"epoch": 0.49,
"grad_norm": 0.39802250266075134,
"learning_rate": 4.291331283908562e-05,
"loss": 1.3269,
"step": 10250
},
{
"epoch": 0.49,
"grad_norm": 0.3359144330024719,
"learning_rate": 4.2900159592259105e-05,
"loss": 1.3396,
"step": 10260
},
{
"epoch": 0.49,
"grad_norm": 0.3711611032485962,
"learning_rate": 4.2886996170036653e-05,
"loss": 1.3383,
"step": 10270
},
{
"epoch": 0.49,
"grad_norm": 0.44124463200569153,
"learning_rate": 4.287382257990105e-05,
"loss": 1.3743,
"step": 10280
},
{
"epoch": 0.49,
"grad_norm": 0.33610624074935913,
"learning_rate": 4.286063882934087e-05,
"loss": 1.3718,
"step": 10290
},
{
"epoch": 0.49,
"grad_norm": 0.369437038898468,
"learning_rate": 4.2847444925850453e-05,
"loss": 1.3751,
"step": 10300
},
{
"epoch": 0.49,
"grad_norm": 0.41726449131965637,
"learning_rate": 4.283424087692991e-05,
"loss": 1.3895,
"step": 10310
},
{
"epoch": 0.5,
"grad_norm": 0.3629494905471802,
"learning_rate": 4.28210266900851e-05,
"loss": 1.3132,
"step": 10320
},
{
"epoch": 0.5,
"grad_norm": 0.400240421295166,
"learning_rate": 4.280780237282769e-05,
"loss": 1.3909,
"step": 10330
},
{
"epoch": 0.5,
"grad_norm": 0.3639288544654846,
"learning_rate": 4.279456793267507e-05,
"loss": 1.3316,
"step": 10340
},
{
"epoch": 0.5,
"grad_norm": 0.32167065143585205,
"learning_rate": 4.278132337715039e-05,
"loss": 1.3109,
"step": 10350
},
{
"epoch": 0.5,
"grad_norm": 0.35733965039253235,
"learning_rate": 4.276806871378256e-05,
"loss": 1.3727,
"step": 10360
},
{
"epoch": 0.5,
"grad_norm": 0.35744282603263855,
"learning_rate": 4.275480395010624e-05,
"loss": 1.3239,
"step": 10370
},
{
"epoch": 0.5,
"grad_norm": 0.3381868302822113,
"learning_rate": 4.2741529093661796e-05,
"loss": 1.4466,
"step": 10380
},
{
"epoch": 0.5,
"grad_norm": 0.36411580443382263,
"learning_rate": 4.272824415199539e-05,
"loss": 1.3806,
"step": 10390
},
{
"epoch": 0.5,
"grad_norm": 0.3639141917228699,
"learning_rate": 4.271494913265885e-05,
"loss": 1.3099,
"step": 10400
},
{
"epoch": 0.5,
"grad_norm": 0.3770304024219513,
"learning_rate": 4.27016440432098e-05,
"loss": 1.38,
"step": 10410
},
{
"epoch": 0.5,
"grad_norm": 0.3780912756919861,
"learning_rate": 4.2688328891211535e-05,
"loss": 1.351,
"step": 10420
},
{
"epoch": 0.5,
"grad_norm": 0.332579642534256,
"learning_rate": 4.2675003684233095e-05,
"loss": 1.3214,
"step": 10430
},
{
"epoch": 0.5,
"grad_norm": 0.3244479298591614,
"learning_rate": 4.2661668429849235e-05,
"loss": 1.3711,
"step": 10440
},
{
"epoch": 0.5,
"grad_norm": 0.3021569550037384,
"learning_rate": 4.264832313564041e-05,
"loss": 1.3355,
"step": 10450
},
{
"epoch": 0.5,
"grad_norm": 0.49049100279808044,
"learning_rate": 4.263496780919279e-05,
"loss": 1.4453,
"step": 10460
},
{
"epoch": 0.5,
"grad_norm": 0.3403632938861847,
"learning_rate": 4.2621602458098265e-05,
"loss": 1.2738,
"step": 10470
},
{
"epoch": 0.5,
"grad_norm": 0.3408461809158325,
"learning_rate": 4.260822708995439e-05,
"loss": 1.3831,
"step": 10480
},
{
"epoch": 0.5,
"grad_norm": 0.3400132358074188,
"learning_rate": 4.259484171236443e-05,
"loss": 1.3768,
"step": 10490
},
{
"epoch": 0.5,
"grad_norm": 0.34560486674308777,
"learning_rate": 4.258144633293735e-05,
"loss": 1.3744,
"step": 10500
},
{
"epoch": 0.5,
"grad_norm": 0.3882208466529846,
"learning_rate": 4.256804095928779e-05,
"loss": 1.3888,
"step": 10510
},
{
"epoch": 0.5,
"grad_norm": 0.3689284920692444,
"learning_rate": 4.2554625599036055e-05,
"loss": 1.3848,
"step": 10520
},
{
"epoch": 0.51,
"grad_norm": 0.4319477677345276,
"learning_rate": 4.2541200259808165e-05,
"loss": 1.3063,
"step": 10530
},
{
"epoch": 0.51,
"grad_norm": 0.47407233715057373,
"learning_rate": 4.2527764949235785e-05,
"loss": 1.3344,
"step": 10540
},
{
"epoch": 0.51,
"grad_norm": 0.40425407886505127,
"learning_rate": 4.251431967495625e-05,
"loss": 1.3744,
"step": 10550
},
{
"epoch": 0.51,
"grad_norm": 0.3393104374408722,
"learning_rate": 4.2500864444612574e-05,
"loss": 1.3647,
"step": 10560
},
{
"epoch": 0.51,
"grad_norm": 0.4788239300251007,
"learning_rate": 4.248739926585341e-05,
"loss": 1.362,
"step": 10570
},
{
"epoch": 0.51,
"grad_norm": 0.36177313327789307,
"learning_rate": 4.2473924146333086e-05,
"loss": 1.3246,
"step": 10580
},
{
"epoch": 0.51,
"grad_norm": 0.3977445662021637,
"learning_rate": 4.2460439093711555e-05,
"loss": 1.4626,
"step": 10590
},
{
"epoch": 0.51,
"grad_norm": 0.38721612095832825,
"learning_rate": 4.244694411565446e-05,
"loss": 1.2976,
"step": 10600
},
{
"epoch": 0.51,
"grad_norm": 0.3681018352508545,
"learning_rate": 4.2433439219833035e-05,
"loss": 1.3912,
"step": 10610
},
{
"epoch": 0.51,
"grad_norm": 0.3572552800178528,
"learning_rate": 4.241992441392418e-05,
"loss": 1.3706,
"step": 10620
},
{
"epoch": 0.51,
"grad_norm": 0.39360326528549194,
"learning_rate": 4.2406399705610434e-05,
"loss": 1.3247,
"step": 10630
},
{
"epoch": 0.51,
"grad_norm": 0.3340846300125122,
"learning_rate": 4.239286510257995e-05,
"loss": 1.354,
"step": 10640
},
{
"epoch": 0.51,
"grad_norm": 0.3777264654636383,
"learning_rate": 4.237932061252651e-05,
"loss": 1.252,
"step": 10650
},
{
"epoch": 0.51,
"grad_norm": 0.3748822510242462,
"learning_rate": 4.236576624314952e-05,
"loss": 1.2653,
"step": 10660
},
{
"epoch": 0.51,
"grad_norm": 0.352718710899353,
"learning_rate": 4.235220200215401e-05,
"loss": 1.357,
"step": 10670
},
{
"epoch": 0.51,
"grad_norm": 0.3392397165298462,
"learning_rate": 4.2338627897250585e-05,
"loss": 1.2837,
"step": 10680
},
{
"epoch": 0.51,
"grad_norm": 0.3369060754776001,
"learning_rate": 4.232504393615549e-05,
"loss": 1.4169,
"step": 10690
},
{
"epoch": 0.51,
"grad_norm": 0.5633871555328369,
"learning_rate": 4.231145012659058e-05,
"loss": 1.333,
"step": 10700
},
{
"epoch": 0.51,
"grad_norm": 0.48663416504859924,
"learning_rate": 4.2297846476283285e-05,
"loss": 1.3861,
"step": 10710
},
{
"epoch": 0.51,
"grad_norm": 0.3928183913230896,
"learning_rate": 4.2284232992966646e-05,
"loss": 1.4284,
"step": 10720
},
{
"epoch": 0.52,
"grad_norm": 0.4091176688671112,
"learning_rate": 4.227060968437928e-05,
"loss": 1.3828,
"step": 10730
},
{
"epoch": 0.52,
"grad_norm": 0.35482415556907654,
"learning_rate": 4.225697655826538e-05,
"loss": 1.3148,
"step": 10740
},
{
"epoch": 0.52,
"grad_norm": 0.4311554431915283,
"learning_rate": 4.224333362237477e-05,
"loss": 1.2308,
"step": 10750
},
{
"epoch": 0.52,
"grad_norm": 0.4161139130592346,
"learning_rate": 4.222968088446278e-05,
"loss": 1.2892,
"step": 10760
},
{
"epoch": 0.52,
"grad_norm": 0.4597773551940918,
"learning_rate": 4.2216018352290374e-05,
"loss": 1.3482,
"step": 10770
},
{
"epoch": 0.52,
"grad_norm": 0.4886460602283478,
"learning_rate": 4.220234603362403e-05,
"loss": 1.3735,
"step": 10780
},
{
"epoch": 0.52,
"grad_norm": 0.33669140934944153,
"learning_rate": 4.218866393623584e-05,
"loss": 1.3616,
"step": 10790
},
{
"epoch": 0.52,
"grad_norm": 0.3581002652645111,
"learning_rate": 4.217497206790342e-05,
"loss": 1.4187,
"step": 10800
},
{
"epoch": 0.52,
"grad_norm": 0.39812520146369934,
"learning_rate": 4.216127043640995e-05,
"loss": 1.3894,
"step": 10810
},
{
"epoch": 0.52,
"grad_norm": 0.302670419216156,
"learning_rate": 4.214755904954416e-05,
"loss": 1.3306,
"step": 10820
},
{
"epoch": 0.52,
"grad_norm": 0.3210192620754242,
"learning_rate": 4.213383791510034e-05,
"loss": 1.3475,
"step": 10830
},
{
"epoch": 0.52,
"grad_norm": 0.5363901853561401,
"learning_rate": 4.212010704087828e-05,
"loss": 1.3302,
"step": 10840
},
{
"epoch": 0.52,
"grad_norm": 0.30793869495391846,
"learning_rate": 4.2106366434683364e-05,
"loss": 1.3919,
"step": 10850
},
{
"epoch": 0.52,
"grad_norm": 0.37882688641548157,
"learning_rate": 4.209261610432646e-05,
"loss": 1.38,
"step": 10860
},
{
"epoch": 0.52,
"grad_norm": 0.374026894569397,
"learning_rate": 4.207885605762399e-05,
"loss": 1.4631,
"step": 10870
},
{
"epoch": 0.52,
"grad_norm": 0.3480188250541687,
"learning_rate": 4.2065086302397884e-05,
"loss": 1.3847,
"step": 10880
},
{
"epoch": 0.52,
"grad_norm": 0.3261585831642151,
"learning_rate": 4.2051306846475604e-05,
"loss": 1.4099,
"step": 10890
},
{
"epoch": 0.52,
"grad_norm": 0.3335872292518616,
"learning_rate": 4.203751769769011e-05,
"loss": 1.3358,
"step": 10900
},
{
"epoch": 0.52,
"grad_norm": 0.3239443898200989,
"learning_rate": 4.20237188638799e-05,
"loss": 1.3744,
"step": 10910
},
{
"epoch": 0.52,
"grad_norm": 0.37974122166633606,
"learning_rate": 4.200991035288895e-05,
"loss": 1.4444,
"step": 10920
},
{
"epoch": 0.52,
"grad_norm": 0.34473395347595215,
"learning_rate": 4.1996092172566735e-05,
"loss": 1.3507,
"step": 10930
},
{
"epoch": 0.53,
"grad_norm": 0.321281373500824,
"learning_rate": 4.1982264330768264e-05,
"loss": 1.3694,
"step": 10940
},
{
"epoch": 0.53,
"grad_norm": 0.3602040410041809,
"learning_rate": 4.196842683535399e-05,
"loss": 1.394,
"step": 10950
},
{
"epoch": 0.53,
"grad_norm": 0.4131835997104645,
"learning_rate": 4.1954579694189886e-05,
"loss": 1.4091,
"step": 10960
},
{
"epoch": 0.53,
"grad_norm": 0.35014450550079346,
"learning_rate": 4.19407229151474e-05,
"loss": 1.3294,
"step": 10970
},
{
"epoch": 0.53,
"grad_norm": 0.3207249641418457,
"learning_rate": 4.192685650610346e-05,
"loss": 1.3369,
"step": 10980
},
{
"epoch": 0.53,
"grad_norm": 0.3043813705444336,
"learning_rate": 4.191298047494045e-05,
"loss": 1.3569,
"step": 10990
},
{
"epoch": 0.53,
"grad_norm": 0.42258018255233765,
"learning_rate": 4.189909482954625e-05,
"loss": 1.3418,
"step": 11000
},
{
"epoch": 0.53,
"grad_norm": 0.5071279406547546,
"learning_rate": 4.188519957781419e-05,
"loss": 1.3698,
"step": 11010
},
{
"epoch": 0.53,
"grad_norm": 0.3340989053249359,
"learning_rate": 4.1871294727643074e-05,
"loss": 1.366,
"step": 11020
},
{
"epoch": 0.53,
"grad_norm": 0.34428247809410095,
"learning_rate": 4.185738028693714e-05,
"loss": 1.3383,
"step": 11030
},
{
"epoch": 0.53,
"grad_norm": 0.3850650191307068,
"learning_rate": 4.184345626360609e-05,
"loss": 1.3351,
"step": 11040
},
{
"epoch": 0.53,
"grad_norm": 0.39173945784568787,
"learning_rate": 4.182952266556509e-05,
"loss": 1.345,
"step": 11050
},
{
"epoch": 0.53,
"grad_norm": 0.31698256731033325,
"learning_rate": 4.1815579500734724e-05,
"loss": 1.3618,
"step": 11060
},
{
"epoch": 0.53,
"grad_norm": 0.42221584916114807,
"learning_rate": 4.180162677704102e-05,
"loss": 1.337,
"step": 11070
},
{
"epoch": 0.53,
"grad_norm": 0.493282288312912,
"learning_rate": 4.178766450241545e-05,
"loss": 1.3491,
"step": 11080
},
{
"epoch": 0.53,
"grad_norm": 0.542514443397522,
"learning_rate": 4.17736926847949e-05,
"loss": 1.3671,
"step": 11090
},
{
"epoch": 0.53,
"grad_norm": 0.4172150492668152,
"learning_rate": 4.1759711332121696e-05,
"loss": 1.4123,
"step": 11100
},
{
"epoch": 0.53,
"grad_norm": 0.36019882559776306,
"learning_rate": 4.174572045234357e-05,
"loss": 1.3181,
"step": 11110
},
{
"epoch": 0.53,
"grad_norm": 0.3032473623752594,
"learning_rate": 4.173172005341369e-05,
"loss": 1.3819,
"step": 11120
},
{
"epoch": 0.53,
"grad_norm": 0.34529924392700195,
"learning_rate": 4.17177101432906e-05,
"loss": 1.3747,
"step": 11130
},
{
"epoch": 0.53,
"grad_norm": 0.5233802795410156,
"learning_rate": 4.17036907299383e-05,
"loss": 1.3809,
"step": 11140
},
{
"epoch": 0.54,
"grad_norm": 0.33382219076156616,
"learning_rate": 4.168966182132613e-05,
"loss": 1.3725,
"step": 11150
},
{
"epoch": 0.54,
"grad_norm": 0.31049731373786926,
"learning_rate": 4.16756234254289e-05,
"loss": 1.3392,
"step": 11160
},
{
"epoch": 0.54,
"grad_norm": 0.4977787137031555,
"learning_rate": 4.1661575550226754e-05,
"loss": 1.3592,
"step": 11170
},
{
"epoch": 0.54,
"grad_norm": 0.34578070044517517,
"learning_rate": 4.164751820370525e-05,
"loss": 1.3351,
"step": 11180
},
{
"epoch": 0.54,
"grad_norm": 0.4598138630390167,
"learning_rate": 4.1633451393855324e-05,
"loss": 1.2932,
"step": 11190
},
{
"epoch": 0.54,
"grad_norm": 0.35738906264305115,
"learning_rate": 4.16193751286733e-05,
"loss": 1.2949,
"step": 11200
},
{
"epoch": 0.54,
"grad_norm": 0.34455692768096924,
"learning_rate": 4.160528941616087e-05,
"loss": 1.3008,
"step": 11210
},
{
"epoch": 0.54,
"grad_norm": 0.3759899437427521,
"learning_rate": 4.159119426432509e-05,
"loss": 1.3,
"step": 11220
},
{
"epoch": 0.54,
"grad_norm": 0.3564906418323517,
"learning_rate": 4.1577089681178396e-05,
"loss": 1.3749,
"step": 11230
},
{
"epoch": 0.54,
"grad_norm": 0.36317357420921326,
"learning_rate": 4.1562975674738565e-05,
"loss": 1.4493,
"step": 11240
},
{
"epoch": 0.54,
"grad_norm": 0.3052905201911926,
"learning_rate": 4.154885225302875e-05,
"loss": 1.407,
"step": 11250
},
{
"epoch": 0.54,
"grad_norm": 0.33644822239875793,
"learning_rate": 4.153471942407746e-05,
"loss": 1.4195,
"step": 11260
},
{
"epoch": 0.54,
"grad_norm": 0.4278157353401184,
"learning_rate": 4.152057719591854e-05,
"loss": 1.3924,
"step": 11270
},
{
"epoch": 0.54,
"grad_norm": 0.4753280580043793,
"learning_rate": 4.150642557659116e-05,
"loss": 1.3341,
"step": 11280
},
{
"epoch": 0.54,
"grad_norm": 0.37113046646118164,
"learning_rate": 4.149226457413986e-05,
"loss": 1.359,
"step": 11290
},
{
"epoch": 0.54,
"grad_norm": 0.409722238779068,
"learning_rate": 4.147809419661451e-05,
"loss": 1.3348,
"step": 11300
},
{
"epoch": 0.54,
"grad_norm": 0.3237214684486389,
"learning_rate": 4.1463914452070286e-05,
"loss": 1.4205,
"step": 11310
},
{
"epoch": 0.54,
"grad_norm": 0.38030579686164856,
"learning_rate": 4.144972534856771e-05,
"loss": 1.3952,
"step": 11320
},
{
"epoch": 0.54,
"grad_norm": 0.34114766120910645,
"learning_rate": 4.1435526894172626e-05,
"loss": 1.3704,
"step": 11330
},
{
"epoch": 0.54,
"grad_norm": 0.34066393971443176,
"learning_rate": 4.1421319096956176e-05,
"loss": 1.3366,
"step": 11340
},
{
"epoch": 0.54,
"grad_norm": 0.457366019487381,
"learning_rate": 4.1407101964994816e-05,
"loss": 1.4163,
"step": 11350
},
{
"epoch": 0.55,
"grad_norm": 0.402567595243454,
"learning_rate": 4.139287550637033e-05,
"loss": 1.2487,
"step": 11360
},
{
"epoch": 0.55,
"grad_norm": 0.4116329848766327,
"learning_rate": 4.137863972916978e-05,
"loss": 1.3735,
"step": 11370
},
{
"epoch": 0.55,
"grad_norm": 0.3859683573246002,
"learning_rate": 4.1364394641485525e-05,
"loss": 1.3031,
"step": 11380
},
{
"epoch": 0.55,
"grad_norm": 0.41960641741752625,
"learning_rate": 4.1350140251415234e-05,
"loss": 1.3013,
"step": 11390
},
{
"epoch": 0.55,
"grad_norm": 0.46500012278556824,
"learning_rate": 4.133587656706186e-05,
"loss": 1.2787,
"step": 11400
},
{
"epoch": 0.55,
"grad_norm": 0.3360355794429779,
"learning_rate": 4.132160359653363e-05,
"loss": 1.4223,
"step": 11410
},
{
"epoch": 0.55,
"grad_norm": 0.4086500108242035,
"learning_rate": 4.130732134794405e-05,
"loss": 1.3897,
"step": 11420
},
{
"epoch": 0.55,
"grad_norm": 0.4114365577697754,
"learning_rate": 4.129302982941191e-05,
"loss": 1.2777,
"step": 11430
},
{
"epoch": 0.55,
"grad_norm": 0.48998185992240906,
"learning_rate": 4.127872904906126e-05,
"loss": 1.308,
"step": 11440
},
{
"epoch": 0.55,
"grad_norm": 0.2914040982723236,
"learning_rate": 4.1264419015021414e-05,
"loss": 1.3472,
"step": 11450
},
{
"epoch": 0.55,
"grad_norm": 0.3826962411403656,
"learning_rate": 4.125009973542697e-05,
"loss": 1.3839,
"step": 11460
},
{
"epoch": 0.55,
"grad_norm": 0.46187523007392883,
"learning_rate": 4.123577121841774e-05,
"loss": 1.2501,
"step": 11470
},
{
"epoch": 0.55,
"grad_norm": 0.4033411145210266,
"learning_rate": 4.122143347213883e-05,
"loss": 1.4376,
"step": 11480
},
{
"epoch": 0.55,
"grad_norm": 0.5302203893661499,
"learning_rate": 4.1207086504740546e-05,
"loss": 1.3614,
"step": 11490
},
{
"epoch": 0.55,
"grad_norm": 0.3035750687122345,
"learning_rate": 4.119273032437849e-05,
"loss": 1.4274,
"step": 11500
},
{
"epoch": 0.55,
"grad_norm": 0.3709616959095001,
"learning_rate": 4.1178364939213456e-05,
"loss": 1.3258,
"step": 11510
},
{
"epoch": 0.55,
"grad_norm": 0.3554280996322632,
"learning_rate": 4.116399035741149e-05,
"loss": 1.352,
"step": 11520
},
{
"epoch": 0.55,
"grad_norm": 0.37478235363960266,
"learning_rate": 4.1149606587143865e-05,
"loss": 1.3703,
"step": 11530
},
{
"epoch": 0.55,
"grad_norm": 0.41338199377059937,
"learning_rate": 4.113521363658708e-05,
"loss": 1.3881,
"step": 11540
},
{
"epoch": 0.55,
"grad_norm": 0.42456531524658203,
"learning_rate": 4.112225213870081e-05,
"loss": 1.3607,
"step": 11550
},
{
"epoch": 0.55,
"grad_norm": 0.36959806084632874,
"learning_rate": 4.110928321737552e-05,
"loss": 1.3307,
"step": 11560
},
{
"epoch": 0.56,
"grad_norm": 0.42485007643699646,
"learning_rate": 4.109486460555243e-05,
"loss": 1.2432,
"step": 11570
},
{
"epoch": 0.56,
"grad_norm": 0.4791394770145416,
"learning_rate": 4.1080436844558404e-05,
"loss": 1.4316,
"step": 11580
},
{
"epoch": 0.56,
"grad_norm": 0.4650508761405945,
"learning_rate": 4.106599994259495e-05,
"loss": 1.2659,
"step": 11590
},
{
"epoch": 0.56,
"grad_norm": 0.345375120639801,
"learning_rate": 4.1051553907868754e-05,
"loss": 1.408,
"step": 11600
},
{
"epoch": 0.56,
"grad_norm": 0.4205334782600403,
"learning_rate": 4.103709874859172e-05,
"loss": 1.3897,
"step": 11610
},
{
"epoch": 0.56,
"grad_norm": 0.3284691274166107,
"learning_rate": 4.102263447298092e-05,
"loss": 1.4017,
"step": 11620
},
{
"epoch": 0.56,
"grad_norm": 0.42488259077072144,
"learning_rate": 4.100816108925861e-05,
"loss": 1.4347,
"step": 11630
},
{
"epoch": 0.56,
"grad_norm": 0.40776416659355164,
"learning_rate": 4.0993678605652234e-05,
"loss": 1.3369,
"step": 11640
},
{
"epoch": 0.56,
"grad_norm": 0.3999396562576294,
"learning_rate": 4.097918703039438e-05,
"loss": 1.3694,
"step": 11650
},
{
"epoch": 0.56,
"grad_norm": 0.37729015946388245,
"learning_rate": 4.096468637172284e-05,
"loss": 1.3447,
"step": 11660
},
{
"epoch": 0.56,
"grad_norm": 0.3748543858528137,
"learning_rate": 4.095017663788056e-05,
"loss": 1.2363,
"step": 11670
},
{
"epoch": 0.56,
"grad_norm": 0.4567815959453583,
"learning_rate": 4.093565783711561e-05,
"loss": 1.3666,
"step": 11680
},
{
"epoch": 0.56,
"grad_norm": 0.4209630787372589,
"learning_rate": 4.0921129977681284e-05,
"loss": 1.422,
"step": 11690
},
{
"epoch": 0.56,
"grad_norm": 0.3567582368850708,
"learning_rate": 4.0906593067835955e-05,
"loss": 1.3543,
"step": 11700
},
{
"epoch": 0.56,
"grad_norm": 0.3691411018371582,
"learning_rate": 4.089204711584317e-05,
"loss": 1.4327,
"step": 11710
},
{
"epoch": 0.56,
"grad_norm": 0.2798110246658325,
"learning_rate": 4.087749212997164e-05,
"loss": 1.3766,
"step": 11720
},
{
"epoch": 0.56,
"grad_norm": 0.4231710731983185,
"learning_rate": 4.086292811849516e-05,
"loss": 1.3714,
"step": 11730
},
{
"epoch": 0.56,
"grad_norm": 0.5270175337791443,
"learning_rate": 4.0848355089692695e-05,
"loss": 1.3162,
"step": 11740
},
{
"epoch": 0.56,
"grad_norm": 0.3277260363101959,
"learning_rate": 4.083377305184833e-05,
"loss": 1.3889,
"step": 11750
},
{
"epoch": 0.56,
"grad_norm": 0.35417479276657104,
"learning_rate": 4.0819182013251256e-05,
"loss": 1.3749,
"step": 11760
},
{
"epoch": 0.56,
"grad_norm": 0.37502023577690125,
"learning_rate": 4.08045819821958e-05,
"loss": 1.3851,
"step": 11770
},
{
"epoch": 0.57,
"grad_norm": 0.42905208468437195,
"learning_rate": 4.078997296698138e-05,
"loss": 1.348,
"step": 11780
},
{
"epoch": 0.57,
"grad_norm": 0.4014531672000885,
"learning_rate": 4.077535497591254e-05,
"loss": 1.2241,
"step": 11790
},
{
"epoch": 0.57,
"grad_norm": 0.334602028131485,
"learning_rate": 4.076072801729891e-05,
"loss": 1.3051,
"step": 11800
},
{
"epoch": 0.57,
"grad_norm": 0.42459970712661743,
"learning_rate": 4.0746092099455225e-05,
"loss": 1.3107,
"step": 11810
},
{
"epoch": 0.57,
"grad_norm": 0.3217390179634094,
"learning_rate": 4.073291212013045e-05,
"loss": 1.3285,
"step": 11820
},
{
"epoch": 0.57,
"grad_norm": 0.40947431325912476,
"learning_rate": 4.0718259202675015e-05,
"loss": 1.3512,
"step": 11830
},
{
"epoch": 0.57,
"grad_norm": 0.36466550827026367,
"learning_rate": 4.0703597350131036e-05,
"loss": 1.3025,
"step": 11840
},
{
"epoch": 0.57,
"grad_norm": 0.34644851088523865,
"learning_rate": 4.0688926570833095e-05,
"loss": 1.2607,
"step": 11850
},
{
"epoch": 0.57,
"grad_norm": 0.34458133578300476,
"learning_rate": 4.0674246873120834e-05,
"loss": 1.3451,
"step": 11860
},
{
"epoch": 0.57,
"grad_norm": 0.354220986366272,
"learning_rate": 4.0659558265338984e-05,
"loss": 1.3773,
"step": 11870
},
{
"epoch": 0.57,
"grad_norm": 0.46973133087158203,
"learning_rate": 4.06448607558373e-05,
"loss": 1.3103,
"step": 11880
},
{
"epoch": 0.57,
"grad_norm": 0.39625704288482666,
"learning_rate": 4.063015435297064e-05,
"loss": 1.3145,
"step": 11890
},
{
"epoch": 0.57,
"grad_norm": 0.7659072875976562,
"learning_rate": 4.06154390650989e-05,
"loss": 1.4046,
"step": 11900
},
{
"epoch": 0.57,
"grad_norm": 0.4457710087299347,
"learning_rate": 4.0600714900587016e-05,
"loss": 1.3902,
"step": 11910
},
{
"epoch": 0.57,
"grad_norm": 0.3246901035308838,
"learning_rate": 4.058598186780499e-05,
"loss": 1.2941,
"step": 11920
},
{
"epoch": 0.57,
"grad_norm": 0.3571087419986725,
"learning_rate": 4.057123997512785e-05,
"loss": 1.3019,
"step": 11930
},
{
"epoch": 0.57,
"grad_norm": 0.5159611105918884,
"learning_rate": 4.0556489230935666e-05,
"loss": 1.3063,
"step": 11940
},
{
"epoch": 0.57,
"grad_norm": 0.3866085410118103,
"learning_rate": 4.0541729643613544e-05,
"loss": 1.3059,
"step": 11950
},
{
"epoch": 0.57,
"grad_norm": 0.3834385573863983,
"learning_rate": 4.052696122155162e-05,
"loss": 1.2822,
"step": 11960
},
{
"epoch": 0.57,
"grad_norm": 0.48380643129348755,
"learning_rate": 4.051218397314502e-05,
"loss": 1.3509,
"step": 11970
},
{
"epoch": 0.58,
"grad_norm": 0.40381157398223877,
"learning_rate": 4.049739790679394e-05,
"loss": 1.3176,
"step": 11980
},
{
"epoch": 0.58,
"grad_norm": 0.3758612275123596,
"learning_rate": 4.0482603030903556e-05,
"loss": 1.2843,
"step": 11990
},
{
"epoch": 0.58,
"grad_norm": 0.36170676350593567,
"learning_rate": 4.046779935388404e-05,
"loss": 1.3408,
"step": 12000
},
{
"epoch": 0.58,
"grad_norm": 0.4872267246246338,
"learning_rate": 4.0452986884150604e-05,
"loss": 1.2936,
"step": 12010
},
{
"epoch": 0.58,
"grad_norm": 0.38034674525260925,
"learning_rate": 4.043816563012341e-05,
"loss": 1.2922,
"step": 12020
},
{
"epoch": 0.58,
"grad_norm": 0.37976744771003723,
"learning_rate": 4.042333560022768e-05,
"loss": 1.3754,
"step": 12030
},
{
"epoch": 0.58,
"grad_norm": 0.3503049314022064,
"learning_rate": 4.0408496802893546e-05,
"loss": 1.3208,
"step": 12040
},
{
"epoch": 0.58,
"grad_norm": 0.3092308044433594,
"learning_rate": 4.03936492465562e-05,
"loss": 1.3426,
"step": 12050
},
{
"epoch": 0.58,
"grad_norm": 0.3568356931209564,
"learning_rate": 4.037879293965575e-05,
"loss": 1.2541,
"step": 12060
},
{
"epoch": 0.58,
"grad_norm": 0.36739981174468994,
"learning_rate": 4.036392789063731e-05,
"loss": 1.3399,
"step": 12070
},
{
"epoch": 0.58,
"grad_norm": 0.34156322479248047,
"learning_rate": 4.034905410795097e-05,
"loss": 1.3447,
"step": 12080
},
{
"epoch": 0.58,
"grad_norm": 0.4482158124446869,
"learning_rate": 4.033417160005177e-05,
"loss": 1.3248,
"step": 12090
},
{
"epoch": 0.58,
"grad_norm": 0.6183753609657288,
"learning_rate": 4.031928037539972e-05,
"loss": 1.2915,
"step": 12100
},
{
"epoch": 0.58,
"grad_norm": 0.37588632106781006,
"learning_rate": 4.030438044245976e-05,
"loss": 1.3494,
"step": 12110
},
{
"epoch": 0.58,
"grad_norm": 0.5009880661964417,
"learning_rate": 4.028947180970181e-05,
"loss": 1.3913,
"step": 12120
},
{
"epoch": 0.58,
"grad_norm": 0.3554256856441498,
"learning_rate": 4.0274554485600726e-05,
"loss": 1.3259,
"step": 12130
},
{
"epoch": 0.58,
"grad_norm": 0.4170083701610565,
"learning_rate": 4.0259628478636295e-05,
"loss": 1.3451,
"step": 12140
},
{
"epoch": 0.58,
"grad_norm": 0.48113930225372314,
"learning_rate": 4.0244693797293254e-05,
"loss": 1.3031,
"step": 12150
},
{
"epoch": 0.58,
"grad_norm": 0.4876077473163605,
"learning_rate": 4.022975045006127e-05,
"loss": 1.4316,
"step": 12160
},
{
"epoch": 0.58,
"grad_norm": 0.33987176418304443,
"learning_rate": 4.02147984454349e-05,
"loss": 1.3147,
"step": 12170
},
{
"epoch": 0.58,
"grad_norm": 0.370390921831131,
"learning_rate": 4.0199837791913697e-05,
"loss": 1.3887,
"step": 12180
},
{
"epoch": 0.59,
"grad_norm": 0.41180726885795593,
"learning_rate": 4.018486849800206e-05,
"loss": 1.372,
"step": 12190
},
{
"epoch": 0.59,
"grad_norm": 0.4270785450935364,
"learning_rate": 4.0169890572209336e-05,
"loss": 1.353,
"step": 12200
},
{
"epoch": 0.59,
"grad_norm": 0.3877605199813843,
"learning_rate": 4.0154904023049754e-05,
"loss": 1.2559,
"step": 12210
},
{
"epoch": 0.59,
"grad_norm": 0.41394832730293274,
"learning_rate": 4.013990885904247e-05,
"loss": 1.4287,
"step": 12220
},
{
"epoch": 0.59,
"grad_norm": 0.3352748453617096,
"learning_rate": 4.012490508871154e-05,
"loss": 1.4038,
"step": 12230
},
{
"epoch": 0.59,
"grad_norm": 0.4093269407749176,
"learning_rate": 4.010989272058588e-05,
"loss": 1.3361,
"step": 12240
},
{
"epoch": 0.59,
"grad_norm": 0.3692108988761902,
"learning_rate": 4.009487176319932e-05,
"loss": 1.2922,
"step": 12250
},
{
"epoch": 0.59,
"grad_norm": 0.3474904000759125,
"learning_rate": 4.0079842225090566e-05,
"loss": 1.3907,
"step": 12260
},
{
"epoch": 0.59,
"grad_norm": 0.4685686528682709,
"learning_rate": 4.0064804114803194e-05,
"loss": 1.3596,
"step": 12270
},
{
"epoch": 0.59,
"grad_norm": 0.46237868070602417,
"learning_rate": 4.004975744088567e-05,
"loss": 1.4021,
"step": 12280
},
{
"epoch": 0.59,
"grad_norm": 0.33251482248306274,
"learning_rate": 4.003470221189132e-05,
"loss": 1.2505,
"step": 12290
},
{
"epoch": 0.59,
"grad_norm": 0.3386325538158417,
"learning_rate": 4.001963843637832e-05,
"loss": 1.3191,
"step": 12300
},
{
"epoch": 0.59,
"grad_norm": 0.4214581847190857,
"learning_rate": 4.000456612290972e-05,
"loss": 1.3655,
"step": 12310
},
{
"epoch": 0.59,
"grad_norm": 0.4238710105419159,
"learning_rate": 3.998948528005342e-05,
"loss": 1.325,
"step": 12320
},
{
"epoch": 0.59,
"grad_norm": 0.4265486001968384,
"learning_rate": 3.9974395916382165e-05,
"loss": 1.2601,
"step": 12330
},
{
"epoch": 0.59,
"grad_norm": 0.4091590642929077,
"learning_rate": 3.995929804047355e-05,
"loss": 1.3795,
"step": 12340
},
{
"epoch": 0.59,
"grad_norm": 0.3130303919315338,
"learning_rate": 3.9944191660910005e-05,
"loss": 1.3465,
"step": 12350
},
{
"epoch": 0.59,
"grad_norm": 0.3626957833766937,
"learning_rate": 3.992907678627878e-05,
"loss": 1.3189,
"step": 12360
},
{
"epoch": 0.59,
"grad_norm": 0.43700531125068665,
"learning_rate": 3.9913953425172e-05,
"loss": 1.3602,
"step": 12370
},
{
"epoch": 0.59,
"grad_norm": 0.42722755670547485,
"learning_rate": 3.9898821586186545e-05,
"loss": 1.3006,
"step": 12380
},
{
"epoch": 0.59,
"grad_norm": 0.3487395942211151,
"learning_rate": 3.988368127792418e-05,
"loss": 1.3974,
"step": 12390
},
{
"epoch": 0.6,
"grad_norm": 0.42503151297569275,
"learning_rate": 3.9868532508991436e-05,
"loss": 1.2761,
"step": 12400
},
{
"epoch": 0.6,
"grad_norm": 0.31663256883621216,
"learning_rate": 3.9853375287999685e-05,
"loss": 1.302,
"step": 12410
},
{
"epoch": 0.6,
"grad_norm": 0.3340023458003998,
"learning_rate": 3.983820962356508e-05,
"loss": 1.268,
"step": 12420
},
{
"epoch": 0.6,
"grad_norm": 0.39232856035232544,
"learning_rate": 3.9823035524308594e-05,
"loss": 1.3681,
"step": 12430
},
{
"epoch": 0.6,
"grad_norm": 0.36313676834106445,
"learning_rate": 3.980785299885599e-05,
"loss": 1.3579,
"step": 12440
},
{
"epoch": 0.6,
"grad_norm": 0.33250030875205994,
"learning_rate": 3.979266205583781e-05,
"loss": 1.3312,
"step": 12450
},
{
"epoch": 0.6,
"grad_norm": 0.4148745536804199,
"learning_rate": 3.97774627038894e-05,
"loss": 1.332,
"step": 12460
},
{
"epoch": 0.6,
"grad_norm": 0.40695691108703613,
"learning_rate": 3.976225495165086e-05,
"loss": 1.4132,
"step": 12470
},
{
"epoch": 0.6,
"grad_norm": 0.385002464056015,
"learning_rate": 3.9747038807767076e-05,
"loss": 1.4112,
"step": 12480
},
{
"epoch": 0.6,
"grad_norm": 0.3813878297805786,
"learning_rate": 3.973181428088772e-05,
"loss": 1.3799,
"step": 12490
},
{
"epoch": 0.6,
"grad_norm": 0.3371710479259491,
"learning_rate": 3.9716581379667224e-05,
"loss": 1.2701,
"step": 12500
},
{
"epoch": 0.6,
"grad_norm": 0.3169586658477783,
"learning_rate": 3.970134011276475e-05,
"loss": 1.3382,
"step": 12510
},
{
"epoch": 0.6,
"grad_norm": 0.42466604709625244,
"learning_rate": 3.9686090488844254e-05,
"loss": 1.3213,
"step": 12520
},
{
"epoch": 0.6,
"grad_norm": 0.3158941864967346,
"learning_rate": 3.967083251657443e-05,
"loss": 1.3745,
"step": 12530
},
{
"epoch": 0.6,
"grad_norm": 0.3632413446903229,
"learning_rate": 3.96555662046287e-05,
"loss": 1.4176,
"step": 12540
},
{
"epoch": 0.6,
"grad_norm": 0.3408544063568115,
"learning_rate": 3.964029156168526e-05,
"loss": 1.288,
"step": 12550
},
{
"epoch": 0.6,
"grad_norm": 0.3705766499042511,
"learning_rate": 3.962500859642702e-05,
"loss": 1.348,
"step": 12560
},
{
"epoch": 0.6,
"grad_norm": 0.31769314408302307,
"learning_rate": 3.9609717317541614e-05,
"loss": 1.3321,
"step": 12570
},
{
"epoch": 0.6,
"grad_norm": 0.38591867685317993,
"learning_rate": 3.9594417733721414e-05,
"loss": 1.3389,
"step": 12580
},
{
"epoch": 0.6,
"grad_norm": 0.48292991518974304,
"learning_rate": 3.957910985366352e-05,
"loss": 1.3582,
"step": 12590
},
{
"epoch": 0.6,
"grad_norm": 0.32158997654914856,
"learning_rate": 3.956379368606974e-05,
"loss": 1.4365,
"step": 12600
},
{
"epoch": 0.61,
"grad_norm": 0.33012765645980835,
"learning_rate": 3.954846923964658e-05,
"loss": 1.3474,
"step": 12610
},
{
"epoch": 0.61,
"grad_norm": 0.36589813232421875,
"learning_rate": 3.9533136523105276e-05,
"loss": 1.3734,
"step": 12620
},
{
"epoch": 0.61,
"grad_norm": 0.325336217880249,
"learning_rate": 3.951779554516175e-05,
"loss": 1.2386,
"step": 12630
},
{
"epoch": 0.61,
"grad_norm": 0.3725907802581787,
"learning_rate": 3.950244631453662e-05,
"loss": 1.3869,
"step": 12640
},
{
"epoch": 0.61,
"grad_norm": 0.4779742658138275,
"learning_rate": 3.948708883995521e-05,
"loss": 1.2969,
"step": 12650
},
{
"epoch": 0.61,
"grad_norm": 0.3245665729045868,
"learning_rate": 3.9471723130147507e-05,
"loss": 1.3452,
"step": 12660
},
{
"epoch": 0.61,
"grad_norm": 0.42773106694221497,
"learning_rate": 3.94563491938482e-05,
"loss": 1.2923,
"step": 12670
},
{
"epoch": 0.61,
"grad_norm": 0.4629528820514679,
"learning_rate": 3.944096703979664e-05,
"loss": 1.3171,
"step": 12680
},
{
"epoch": 0.61,
"grad_norm": 0.3126998543739319,
"learning_rate": 3.942557667673686e-05,
"loss": 1.3435,
"step": 12690
},
{
"epoch": 0.61,
"grad_norm": 0.4204506576061249,
"learning_rate": 3.941017811341755e-05,
"loss": 1.3452,
"step": 12700
},
{
"epoch": 0.61,
"grad_norm": 0.4179898798465729,
"learning_rate": 3.939477135859209e-05,
"loss": 1.3515,
"step": 12710
},
{
"epoch": 0.61,
"grad_norm": 0.32224392890930176,
"learning_rate": 3.9379356421018456e-05,
"loss": 1.3088,
"step": 12720
},
{
"epoch": 0.61,
"grad_norm": 0.3186015188694,
"learning_rate": 3.936393330945933e-05,
"loss": 1.425,
"step": 12730
},
{
"epoch": 0.61,
"grad_norm": 0.36656272411346436,
"learning_rate": 3.9348502032682024e-05,
"loss": 1.4293,
"step": 12740
},
{
"epoch": 0.61,
"grad_norm": 0.33963748812675476,
"learning_rate": 3.93330625994585e-05,
"loss": 1.3625,
"step": 12750
},
{
"epoch": 0.61,
"grad_norm": 0.4174005091190338,
"learning_rate": 3.931761501856533e-05,
"loss": 1.3741,
"step": 12760
},
{
"epoch": 0.61,
"grad_norm": 0.3475840389728546,
"learning_rate": 3.9302159298783736e-05,
"loss": 1.3087,
"step": 12770
},
{
"epoch": 0.61,
"grad_norm": 0.35422006249427795,
"learning_rate": 3.928669544889958e-05,
"loss": 1.3317,
"step": 12780
},
{
"epoch": 0.61,
"grad_norm": 0.38269224762916565,
"learning_rate": 3.927122347770332e-05,
"loss": 1.3482,
"step": 12790
},
{
"epoch": 0.61,
"grad_norm": 0.3389624059200287,
"learning_rate": 3.925574339399004e-05,
"loss": 1.3071,
"step": 12800
},
{
"epoch": 0.61,
"grad_norm": 0.37614402174949646,
"learning_rate": 3.924025520655944e-05,
"loss": 1.3493,
"step": 12810
},
{
"epoch": 0.62,
"grad_norm": 0.3476170003414154,
"learning_rate": 3.922475892421582e-05,
"loss": 1.3704,
"step": 12820
},
{
"epoch": 0.62,
"grad_norm": 0.3598422706127167,
"learning_rate": 3.92092545557681e-05,
"loss": 1.3603,
"step": 12830
},
{
"epoch": 0.62,
"grad_norm": 0.36903733015060425,
"learning_rate": 3.919374211002976e-05,
"loss": 1.3269,
"step": 12840
},
{
"epoch": 0.62,
"grad_norm": 0.3061486780643463,
"learning_rate": 3.9178221595818914e-05,
"loss": 1.3179,
"step": 12850
},
{
"epoch": 0.62,
"grad_norm": 0.39297470450401306,
"learning_rate": 3.916269302195823e-05,
"loss": 1.3178,
"step": 12860
},
{
"epoch": 0.62,
"grad_norm": 0.39693140983581543,
"learning_rate": 3.914715639727496e-05,
"loss": 1.3815,
"step": 12870
},
{
"epoch": 0.62,
"grad_norm": 0.4705784320831299,
"learning_rate": 3.9131611730600964e-05,
"loss": 1.349,
"step": 12880
},
{
"epoch": 0.62,
"grad_norm": 0.3764297068119049,
"learning_rate": 3.911605903077264e-05,
"loss": 1.2975,
"step": 12890
},
{
"epoch": 0.62,
"grad_norm": 0.41002950072288513,
"learning_rate": 3.910049830663096e-05,
"loss": 1.3361,
"step": 12900
},
{
"epoch": 0.62,
"grad_norm": 0.40143468976020813,
"learning_rate": 3.908492956702147e-05,
"loss": 1.3862,
"step": 12910
},
{
"epoch": 0.62,
"grad_norm": 0.3195609152317047,
"learning_rate": 3.906935282079425e-05,
"loss": 1.3779,
"step": 12920
},
{
"epoch": 0.62,
"grad_norm": 0.34867754578590393,
"learning_rate": 3.905376807680396e-05,
"loss": 1.3538,
"step": 12930
},
{
"epoch": 0.62,
"grad_norm": 0.3462074398994446,
"learning_rate": 3.903817534390979e-05,
"loss": 1.2739,
"step": 12940
},
{
"epoch": 0.62,
"grad_norm": 0.32215580344200134,
"learning_rate": 3.9022574630975466e-05,
"loss": 1.3543,
"step": 12950
},
{
"epoch": 0.62,
"grad_norm": 0.40852466225624084,
"learning_rate": 3.9006965946869263e-05,
"loss": 1.3007,
"step": 12960
},
{
"epoch": 0.62,
"grad_norm": 0.3892562687397003,
"learning_rate": 3.899134930046397e-05,
"loss": 1.3293,
"step": 12970
},
{
"epoch": 0.62,
"grad_norm": 0.34782683849334717,
"learning_rate": 3.8975724700636915e-05,
"loss": 1.4498,
"step": 12980
},
{
"epoch": 0.62,
"grad_norm": 0.3590206503868103,
"learning_rate": 3.896009215626997e-05,
"loss": 1.2941,
"step": 12990
},
{
"epoch": 0.62,
"grad_norm": 0.44646164774894714,
"learning_rate": 3.8944451676249474e-05,
"loss": 1.4014,
"step": 13000
},
{
"epoch": 0.62,
"grad_norm": 0.33853819966316223,
"learning_rate": 3.892880326946631e-05,
"loss": 1.3452,
"step": 13010
},
{
"epoch": 0.62,
"grad_norm": 0.405666321516037,
"learning_rate": 3.891314694481585e-05,
"loss": 1.3252,
"step": 13020
},
{
"epoch": 0.63,
"grad_norm": 0.3521741032600403,
"learning_rate": 3.889748271119798e-05,
"loss": 1.2438,
"step": 13030
},
{
"epoch": 0.63,
"grad_norm": 0.4369531571865082,
"learning_rate": 3.8881810577517084e-05,
"loss": 1.3595,
"step": 13040
},
{
"epoch": 0.63,
"grad_norm": 0.3240819573402405,
"learning_rate": 3.886613055268203e-05,
"loss": 1.2613,
"step": 13050
},
{
"epoch": 0.63,
"grad_norm": 0.38471776247024536,
"learning_rate": 3.885044264560618e-05,
"loss": 1.3592,
"step": 13060
},
{
"epoch": 0.63,
"grad_norm": 0.4194634258747101,
"learning_rate": 3.883474686520734e-05,
"loss": 1.3833,
"step": 13070
},
{
"epoch": 0.63,
"grad_norm": 0.5312750339508057,
"learning_rate": 3.881904322040784e-05,
"loss": 1.4249,
"step": 13080
},
{
"epoch": 0.63,
"grad_norm": 0.36652135848999023,
"learning_rate": 3.880333172013446e-05,
"loss": 1.3456,
"step": 13090
},
{
"epoch": 0.63,
"grad_norm": 0.4704952836036682,
"learning_rate": 3.878761237331845e-05,
"loss": 1.3047,
"step": 13100
},
{
"epoch": 0.63,
"grad_norm": 0.34810978174209595,
"learning_rate": 3.877188518889551e-05,
"loss": 1.287,
"step": 13110
},
{
"epoch": 0.63,
"grad_norm": 0.32425567507743835,
"learning_rate": 3.875615017580579e-05,
"loss": 1.32,
"step": 13120
},
{
"epoch": 0.63,
"grad_norm": 0.4122347831726074,
"learning_rate": 3.874040734299394e-05,
"loss": 1.3241,
"step": 13130
},
{
"epoch": 0.63,
"grad_norm": 0.35432571172714233,
"learning_rate": 3.872465669940898e-05,
"loss": 1.3341,
"step": 13140
},
{
"epoch": 0.63,
"grad_norm": 0.42856234312057495,
"learning_rate": 3.8708898254004424e-05,
"loss": 1.306,
"step": 13150
},
{
"epoch": 0.63,
"grad_norm": 0.35589203238487244,
"learning_rate": 3.869470898998821e-05,
"loss": 1.3001,
"step": 13160
},
{
"epoch": 0.63,
"grad_norm": 0.3454817235469818,
"learning_rate": 3.8678935745809185e-05,
"loss": 1.3889,
"step": 13170
},
{
"epoch": 0.63,
"grad_norm": 0.3964018225669861,
"learning_rate": 3.8663154725800754e-05,
"loss": 1.2072,
"step": 13180
},
{
"epoch": 0.63,
"grad_norm": 0.3457452654838562,
"learning_rate": 3.864736593893371e-05,
"loss": 1.4434,
"step": 13190
},
{
"epoch": 0.63,
"grad_norm": 0.33705487847328186,
"learning_rate": 3.863156939418321e-05,
"loss": 1.3634,
"step": 13200
},
{
"epoch": 0.63,
"grad_norm": 0.3807193338871002,
"learning_rate": 3.861576510052885e-05,
"loss": 1.3824,
"step": 13210
},
{
"epoch": 0.63,
"grad_norm": 0.3470761477947235,
"learning_rate": 3.859995306695463e-05,
"loss": 1.3936,
"step": 13220
},
{
"epoch": 0.64,
"grad_norm": 0.36898273229599,
"learning_rate": 3.8584133302448935e-05,
"loss": 1.3136,
"step": 13230
},
{
"epoch": 0.64,
"grad_norm": 0.35723429918289185,
"learning_rate": 3.856830581600457e-05,
"loss": 1.3025,
"step": 13240
},
{
"epoch": 0.64,
"grad_norm": 0.34691134095191956,
"learning_rate": 3.85524706166187e-05,
"loss": 1.3547,
"step": 13250
},
{
"epoch": 0.64,
"grad_norm": 0.31771349906921387,
"learning_rate": 3.853662771329288e-05,
"loss": 1.3707,
"step": 13260
},
{
"epoch": 0.64,
"grad_norm": 0.40151447057724,
"learning_rate": 3.852077711503309e-05,
"loss": 1.3196,
"step": 13270
},
{
"epoch": 0.64,
"grad_norm": 0.5305848717689514,
"learning_rate": 3.8504918830849604e-05,
"loss": 1.3376,
"step": 13280
},
{
"epoch": 0.64,
"grad_norm": 0.30876395106315613,
"learning_rate": 3.848905286975714e-05,
"loss": 1.3403,
"step": 13290
},
{
"epoch": 0.64,
"grad_norm": 0.31486690044403076,
"learning_rate": 3.8473179240774725e-05,
"loss": 1.3117,
"step": 13300
},
{
"epoch": 0.64,
"grad_norm": 0.41357141733169556,
"learning_rate": 3.845729795292579e-05,
"loss": 1.245,
"step": 13310
},
{
"epoch": 0.64,
"grad_norm": 0.34300822019577026,
"learning_rate": 3.8441409015238074e-05,
"loss": 1.3363,
"step": 13320
},
{
"epoch": 0.64,
"grad_norm": 0.3271210193634033,
"learning_rate": 3.84255124367437e-05,
"loss": 1.3263,
"step": 13330
},
{
"epoch": 0.64,
"grad_norm": 0.6825655102729797,
"learning_rate": 3.840960822647912e-05,
"loss": 1.4346,
"step": 13340
},
{
"epoch": 0.64,
"grad_norm": 0.40620753169059753,
"learning_rate": 3.8393696393485145e-05,
"loss": 1.3537,
"step": 13350
},
{
"epoch": 0.64,
"grad_norm": 0.39441612362861633,
"learning_rate": 3.8377776946806874e-05,
"loss": 1.4042,
"step": 13360
},
{
"epoch": 0.64,
"grad_norm": 0.34983253479003906,
"learning_rate": 3.836184989549377e-05,
"loss": 1.3415,
"step": 13370
},
{
"epoch": 0.64,
"grad_norm": 0.40179505944252014,
"learning_rate": 3.8345915248599626e-05,
"loss": 1.3353,
"step": 13380
},
{
"epoch": 0.64,
"grad_norm": 0.41166236996650696,
"learning_rate": 3.832997301518251e-05,
"loss": 1.3373,
"step": 13390
},
{
"epoch": 0.64,
"grad_norm": 0.3439750373363495,
"learning_rate": 3.831402320430486e-05,
"loss": 1.3064,
"step": 13400
},
{
"epoch": 0.64,
"grad_norm": 0.3932168185710907,
"learning_rate": 3.829966190327974e-05,
"loss": 1.3565,
"step": 13410
},
{
"epoch": 0.64,
"grad_norm": 0.5207496881484985,
"learning_rate": 3.828369772020939e-05,
"loss": 1.3826,
"step": 13420
},
{
"epoch": 0.64,
"grad_norm": 0.3172326683998108,
"learning_rate": 3.8267725985983804e-05,
"loss": 1.3397,
"step": 13430
},
{
"epoch": 0.65,
"grad_norm": 0.4609996974468231,
"learning_rate": 3.825174670968217e-05,
"loss": 1.3881,
"step": 13440
},
{
"epoch": 0.65,
"grad_norm": 0.5876955389976501,
"learning_rate": 3.823575990038794e-05,
"loss": 1.3001,
"step": 13450
},
{
"epoch": 0.65,
"grad_norm": 0.4091070890426636,
"learning_rate": 3.821976556718888e-05,
"loss": 1.3282,
"step": 13460
},
{
"epoch": 0.65,
"grad_norm": 0.3420022428035736,
"learning_rate": 3.820376371917699e-05,
"loss": 1.2981,
"step": 13470
},
{
"epoch": 0.65,
"grad_norm": 0.3858625888824463,
"learning_rate": 3.818775436544859e-05,
"loss": 1.2607,
"step": 13480
},
{
"epoch": 0.65,
"grad_norm": 0.35949644446372986,
"learning_rate": 3.817173751510421e-05,
"loss": 1.356,
"step": 13490
},
{
"epoch": 0.65,
"grad_norm": 0.4184299409389496,
"learning_rate": 3.815571317724871e-05,
"loss": 1.2999,
"step": 13500
},
{
"epoch": 0.65,
"grad_norm": 0.3372385501861572,
"learning_rate": 3.813968136099115e-05,
"loss": 1.3411,
"step": 13510
},
{
"epoch": 0.65,
"grad_norm": 0.3190743327140808,
"learning_rate": 3.8123642075444856e-05,
"loss": 1.3776,
"step": 13520
},
{
"epoch": 0.65,
"grad_norm": 0.30968132615089417,
"learning_rate": 3.8107595329727416e-05,
"loss": 1.2862,
"step": 13530
},
{
"epoch": 0.65,
"grad_norm": 0.4910031259059906,
"learning_rate": 3.8091541132960644e-05,
"loss": 1.4937,
"step": 13540
},
{
"epoch": 0.65,
"grad_norm": 0.32637661695480347,
"learning_rate": 3.8075479494270593e-05,
"loss": 1.3669,
"step": 13550
},
{
"epoch": 0.65,
"grad_norm": 0.46127691864967346,
"learning_rate": 3.805941042278755e-05,
"loss": 1.3799,
"step": 13560
},
{
"epoch": 0.65,
"grad_norm": 0.5415513515472412,
"learning_rate": 3.804333392764602e-05,
"loss": 1.4008,
"step": 13570
},
{
"epoch": 0.65,
"grad_norm": 0.4115656614303589,
"learning_rate": 3.802725001798473e-05,
"loss": 1.3685,
"step": 13580
},
{
"epoch": 0.65,
"grad_norm": 0.2920158803462982,
"learning_rate": 3.801115870294662e-05,
"loss": 1.3157,
"step": 13590
},
{
"epoch": 0.65,
"grad_norm": 0.32957080006599426,
"learning_rate": 3.799505999167885e-05,
"loss": 1.314,
"step": 13600
},
{
"epoch": 0.65,
"grad_norm": 0.39769747853279114,
"learning_rate": 3.7978953893332786e-05,
"loss": 1.3928,
"step": 13610
},
{
"epoch": 0.65,
"grad_norm": 0.47987455129623413,
"learning_rate": 3.796284041706397e-05,
"loss": 1.3623,
"step": 13620
},
{
"epoch": 0.65,
"grad_norm": 0.3641687035560608,
"learning_rate": 3.794671957203216e-05,
"loss": 1.22,
"step": 13630
},
{
"epoch": 0.65,
"grad_norm": 0.41465121507644653,
"learning_rate": 3.793059136740129e-05,
"loss": 1.3677,
"step": 13640
},
{
"epoch": 0.66,
"grad_norm": 0.32239463925361633,
"learning_rate": 3.791445581233949e-05,
"loss": 1.3734,
"step": 13650
},
{
"epoch": 0.66,
"grad_norm": 0.45050951838493347,
"learning_rate": 3.789831291601905e-05,
"loss": 1.3623,
"step": 13660
},
{
"epoch": 0.66,
"grad_norm": 0.3201011121273041,
"learning_rate": 3.7882162687616465e-05,
"loss": 1.254,
"step": 13670
},
{
"epoch": 0.66,
"grad_norm": 0.36789342761039734,
"learning_rate": 3.786600513631236e-05,
"loss": 1.3292,
"step": 13680
},
{
"epoch": 0.66,
"grad_norm": 0.35849452018737793,
"learning_rate": 3.784984027129154e-05,
"loss": 1.401,
"step": 13690
},
{
"epoch": 0.66,
"grad_norm": 0.33318138122558594,
"learning_rate": 3.783366810174298e-05,
"loss": 1.3394,
"step": 13700
},
{
"epoch": 0.66,
"grad_norm": 0.46340060234069824,
"learning_rate": 3.7817488636859796e-05,
"loss": 1.3714,
"step": 13710
},
{
"epoch": 0.66,
"grad_norm": 0.29935428500175476,
"learning_rate": 3.7801301885839236e-05,
"loss": 1.3878,
"step": 13720
},
{
"epoch": 0.66,
"grad_norm": 0.42615559697151184,
"learning_rate": 3.7785107857882705e-05,
"loss": 1.3763,
"step": 13730
},
{
"epoch": 0.66,
"grad_norm": 0.3408364951610565,
"learning_rate": 3.776890656219576e-05,
"loss": 1.2895,
"step": 13740
},
{
"epoch": 0.66,
"grad_norm": 0.4254874885082245,
"learning_rate": 3.775269800798806e-05,
"loss": 1.388,
"step": 13750
},
{
"epoch": 0.66,
"grad_norm": 0.4857456684112549,
"learning_rate": 3.77364822044734e-05,
"loss": 1.3024,
"step": 13760
},
{
"epoch": 0.66,
"grad_norm": 0.4740150570869446,
"learning_rate": 3.772025916086971e-05,
"loss": 1.3192,
"step": 13770
},
{
"epoch": 0.66,
"grad_norm": 0.34234899282455444,
"learning_rate": 3.770402888639902e-05,
"loss": 1.3209,
"step": 13780
},
{
"epoch": 0.66,
"grad_norm": 0.5989863276481628,
"learning_rate": 3.768779139028748e-05,
"loss": 1.3365,
"step": 13790
},
{
"epoch": 0.66,
"grad_norm": 0.36366432905197144,
"learning_rate": 3.7671546681765326e-05,
"loss": 1.3734,
"step": 13800
},
{
"epoch": 0.66,
"grad_norm": 0.42132848501205444,
"learning_rate": 3.765529477006692e-05,
"loss": 1.3773,
"step": 13810
},
{
"epoch": 0.66,
"grad_norm": 0.42550191283226013,
"learning_rate": 3.76390356644307e-05,
"loss": 1.3918,
"step": 13820
},
{
"epoch": 0.66,
"grad_norm": 0.34290602803230286,
"learning_rate": 3.7622769374099206e-05,
"loss": 1.3876,
"step": 13830
},
{
"epoch": 0.66,
"grad_norm": 0.4383591413497925,
"learning_rate": 3.7606495908319053e-05,
"loss": 1.3726,
"step": 13840
},
{
"epoch": 0.66,
"grad_norm": 0.43107420206069946,
"learning_rate": 3.759021527634095e-05,
"loss": 1.4436,
"step": 13850
},
{
"epoch": 0.67,
"grad_norm": 0.39942875504493713,
"learning_rate": 3.7573927487419644e-05,
"loss": 1.3667,
"step": 13860
},
{
"epoch": 0.67,
"grad_norm": 0.3520909547805786,
"learning_rate": 3.7557632550813994e-05,
"loss": 1.2585,
"step": 13870
},
{
"epoch": 0.67,
"grad_norm": 0.3367423713207245,
"learning_rate": 3.75413304757869e-05,
"loss": 1.2571,
"step": 13880
},
{
"epoch": 0.67,
"grad_norm": 0.34939172863960266,
"learning_rate": 3.7525021271605304e-05,
"loss": 1.3723,
"step": 13890
},
{
"epoch": 0.67,
"grad_norm": 0.35476061701774597,
"learning_rate": 3.7508704947540244e-05,
"loss": 1.273,
"step": 13900
},
{
"epoch": 0.67,
"grad_norm": 0.3022565245628357,
"learning_rate": 3.749238151286677e-05,
"loss": 1.3699,
"step": 13910
},
{
"epoch": 0.67,
"grad_norm": 0.38868963718414307,
"learning_rate": 3.747605097686398e-05,
"loss": 1.3184,
"step": 13920
},
{
"epoch": 0.67,
"grad_norm": 0.6100253462791443,
"learning_rate": 3.7459713348815024e-05,
"loss": 1.3504,
"step": 13930
},
{
"epoch": 0.67,
"grad_norm": 0.3183211386203766,
"learning_rate": 3.7443368638007056e-05,
"loss": 1.3464,
"step": 13940
},
{
"epoch": 0.67,
"grad_norm": 0.5288223624229431,
"learning_rate": 3.742701685373129e-05,
"loss": 1.3822,
"step": 13950
},
{
"epoch": 0.67,
"grad_norm": 0.3603142201900482,
"learning_rate": 3.741065800528293e-05,
"loss": 1.3872,
"step": 13960
},
{
"epoch": 0.67,
"grad_norm": 0.42883366346359253,
"learning_rate": 3.739429210196121e-05,
"loss": 1.3037,
"step": 13970
},
{
"epoch": 0.67,
"grad_norm": 0.40018734335899353,
"learning_rate": 3.7377919153069396e-05,
"loss": 1.3809,
"step": 13980
},
{
"epoch": 0.67,
"grad_norm": 0.5103595852851868,
"learning_rate": 3.73615391679147e-05,
"loss": 1.3012,
"step": 13990
},
{
"epoch": 0.67,
"grad_norm": 0.42257046699523926,
"learning_rate": 3.73451521558084e-05,
"loss": 1.3435,
"step": 14000
},
{
"epoch": 0.67,
"grad_norm": 0.372916042804718,
"learning_rate": 3.732875812606573e-05,
"loss": 1.3228,
"step": 14010
},
{
"epoch": 0.67,
"grad_norm": 0.38566532731056213,
"learning_rate": 3.731235708800592e-05,
"loss": 1.3936,
"step": 14020
},
{
"epoch": 0.67,
"grad_norm": 0.4449050724506378,
"learning_rate": 3.729594905095219e-05,
"loss": 1.3106,
"step": 14030
},
{
"epoch": 0.67,
"grad_norm": 0.43648314476013184,
"learning_rate": 3.7279534024231716e-05,
"loss": 1.3222,
"step": 14040
},
{
"epoch": 0.67,
"grad_norm": 0.5554038286209106,
"learning_rate": 3.726311201717569e-05,
"loss": 1.3422,
"step": 14050
},
{
"epoch": 0.67,
"grad_norm": 0.4891124367713928,
"learning_rate": 3.7246683039119245e-05,
"loss": 1.3565,
"step": 14060
},
{
"epoch": 0.68,
"grad_norm": 0.41038385033607483,
"learning_rate": 3.7230247099401475e-05,
"loss": 1.3611,
"step": 14070
},
{
"epoch": 0.68,
"grad_norm": 0.4609294533729553,
"learning_rate": 3.721380420736543e-05,
"loss": 1.2971,
"step": 14080
},
{
"epoch": 0.68,
"grad_norm": 0.3758825659751892,
"learning_rate": 3.7197354372358114e-05,
"loss": 1.3058,
"step": 14090
},
{
"epoch": 0.68,
"grad_norm": 0.4859221279621124,
"learning_rate": 3.718089760373051e-05,
"loss": 1.412,
"step": 14100
},
{
"epoch": 0.68,
"grad_norm": 0.4632900059223175,
"learning_rate": 3.7164433910837496e-05,
"loss": 1.3258,
"step": 14110
},
{
"epoch": 0.68,
"grad_norm": 0.34527719020843506,
"learning_rate": 3.714796330303789e-05,
"loss": 1.274,
"step": 14120
},
{
"epoch": 0.68,
"grad_norm": 0.5124931931495667,
"learning_rate": 3.7131485789694476e-05,
"loss": 1.3,
"step": 14130
},
{
"epoch": 0.68,
"grad_norm": 0.31964927911758423,
"learning_rate": 3.711500138017393e-05,
"loss": 1.3368,
"step": 14140
},
{
"epoch": 0.68,
"grad_norm": 0.4019540548324585,
"learning_rate": 3.7098510083846875e-05,
"loss": 1.3226,
"step": 14150
},
{
"epoch": 0.68,
"grad_norm": 0.3679623007774353,
"learning_rate": 3.708201191008782e-05,
"loss": 1.294,
"step": 14160
},
{
"epoch": 0.68,
"grad_norm": 0.3462458848953247,
"learning_rate": 3.7065506868275215e-05,
"loss": 1.3113,
"step": 14170
},
{
"epoch": 0.68,
"grad_norm": 0.5903054475784302,
"learning_rate": 3.704899496779138e-05,
"loss": 1.2854,
"step": 14180
},
{
"epoch": 0.68,
"grad_norm": 0.34054383635520935,
"learning_rate": 3.703247621802255e-05,
"loss": 1.359,
"step": 14190
},
{
"epoch": 0.68,
"grad_norm": 0.43107184767723083,
"learning_rate": 3.701595062835887e-05,
"loss": 1.3405,
"step": 14200
},
{
"epoch": 0.68,
"grad_norm": 0.3097091019153595,
"learning_rate": 3.6999418208194355e-05,
"loss": 1.3379,
"step": 14210
},
{
"epoch": 0.68,
"grad_norm": 0.40961962938308716,
"learning_rate": 3.698287896692689e-05,
"loss": 1.3962,
"step": 14220
},
{
"epoch": 0.68,
"grad_norm": 0.3606812059879303,
"learning_rate": 3.696633291395828e-05,
"loss": 1.3004,
"step": 14230
},
{
"epoch": 0.68,
"grad_norm": 0.42880964279174805,
"learning_rate": 3.694978005869415e-05,
"loss": 1.3253,
"step": 14240
},
{
"epoch": 0.68,
"grad_norm": 0.3437255024909973,
"learning_rate": 3.6933220410544024e-05,
"loss": 1.3083,
"step": 14250
},
{
"epoch": 0.68,
"grad_norm": 0.41061654686927795,
"learning_rate": 3.6916653978921286e-05,
"loss": 1.3396,
"step": 14260
},
{
"epoch": 0.68,
"grad_norm": 0.3296366333961487,
"learning_rate": 3.6900080773243165e-05,
"loss": 1.2635,
"step": 14270
},
{
"epoch": 0.69,
"grad_norm": 0.3294108211994171,
"learning_rate": 3.6883500802930754e-05,
"loss": 1.2686,
"step": 14280
},
{
"epoch": 0.69,
"grad_norm": 0.4102252125740051,
"learning_rate": 3.686691407740897e-05,
"loss": 1.3761,
"step": 14290
},
{
"epoch": 0.69,
"grad_norm": 0.34468889236450195,
"learning_rate": 3.68503206061066e-05,
"loss": 1.3519,
"step": 14300
},
{
"epoch": 0.69,
"grad_norm": 0.38928839564323425,
"learning_rate": 3.683372039845623e-05,
"loss": 1.3498,
"step": 14310
},
{
"epoch": 0.69,
"grad_norm": 0.45969000458717346,
"learning_rate": 3.6817113463894317e-05,
"loss": 1.2771,
"step": 14320
},
{
"epoch": 0.69,
"grad_norm": 0.32401660084724426,
"learning_rate": 3.680049981186111e-05,
"loss": 1.3063,
"step": 14330
},
{
"epoch": 0.69,
"grad_norm": 0.3315308094024658,
"learning_rate": 3.6783879451800674e-05,
"loss": 1.391,
"step": 14340
},
{
"epoch": 0.69,
"grad_norm": 0.3669288754463196,
"learning_rate": 3.676725239316092e-05,
"loss": 1.3702,
"step": 14350
},
{
"epoch": 0.69,
"grad_norm": 0.46979033946990967,
"learning_rate": 3.6750618645393525e-05,
"loss": 1.3417,
"step": 14360
},
{
"epoch": 0.69,
"grad_norm": 0.30862849950790405,
"learning_rate": 3.673397821795401e-05,
"loss": 1.3907,
"step": 14370
},
{
"epoch": 0.69,
"grad_norm": 0.4149013161659241,
"learning_rate": 3.6717331120301657e-05,
"loss": 1.3589,
"step": 14380
},
{
"epoch": 0.69,
"grad_norm": 0.42339596152305603,
"learning_rate": 3.670067736189956e-05,
"loss": 1.3859,
"step": 14390
},
{
"epoch": 0.69,
"grad_norm": 0.3184184730052948,
"learning_rate": 3.6684016952214605e-05,
"loss": 1.3743,
"step": 14400
},
{
"epoch": 0.69,
"grad_norm": 0.36168068647384644,
"learning_rate": 3.666734990071743e-05,
"loss": 1.3232,
"step": 14410
},
{
"epoch": 0.69,
"grad_norm": 0.3491245210170746,
"learning_rate": 3.665067621688249e-05,
"loss": 1.3291,
"step": 14420
},
{
"epoch": 0.69,
"grad_norm": 0.36783841252326965,
"learning_rate": 3.663399591018796e-05,
"loss": 1.2747,
"step": 14430
},
{
"epoch": 0.69,
"grad_norm": 0.549431324005127,
"learning_rate": 3.661730899011582e-05,
"loss": 1.4153,
"step": 14440
},
{
"epoch": 0.69,
"grad_norm": 0.4304824471473694,
"learning_rate": 3.6600615466151814e-05,
"loss": 1.3195,
"step": 14450
},
{
"epoch": 0.69,
"grad_norm": 0.4841570258140564,
"learning_rate": 3.6583915347785394e-05,
"loss": 1.2968,
"step": 14460
},
{
"epoch": 0.69,
"grad_norm": 0.36691781878471375,
"learning_rate": 3.65672086445098e-05,
"loss": 1.4298,
"step": 14470
},
{
"epoch": 0.7,
"grad_norm": 0.3750869631767273,
"learning_rate": 3.655049536582202e-05,
"loss": 1.4749,
"step": 14480
},
{
"epoch": 0.7,
"grad_norm": 0.3293134868144989,
"learning_rate": 3.653377552122274e-05,
"loss": 1.1961,
"step": 14490
},
{
"epoch": 0.7,
"grad_norm": 0.3900418281555176,
"learning_rate": 3.651704912021641e-05,
"loss": 1.415,
"step": 14500
},
{
"epoch": 0.7,
"grad_norm": 0.510381817817688,
"learning_rate": 3.6500316172311205e-05,
"loss": 1.3491,
"step": 14510
},
{
"epoch": 0.7,
"grad_norm": 0.4898764491081238,
"learning_rate": 3.648357668701901e-05,
"loss": 1.2808,
"step": 14520
},
{
"epoch": 0.7,
"grad_norm": 0.36868011951446533,
"learning_rate": 3.6466830673855445e-05,
"loss": 1.3871,
"step": 14530
},
{
"epoch": 0.7,
"grad_norm": 0.6004432439804077,
"learning_rate": 3.6450078142339806e-05,
"loss": 1.3683,
"step": 14540
},
{
"epoch": 0.7,
"grad_norm": 0.3484504818916321,
"learning_rate": 3.643331910199515e-05,
"loss": 1.414,
"step": 14550
},
{
"epoch": 0.7,
"grad_norm": 0.3334188163280487,
"learning_rate": 3.641655356234815e-05,
"loss": 1.205,
"step": 14560
},
{
"epoch": 0.7,
"grad_norm": 0.48945480585098267,
"learning_rate": 3.6399781532929276e-05,
"loss": 1.3608,
"step": 14570
},
{
"epoch": 0.7,
"grad_norm": 0.29439377784729004,
"learning_rate": 3.638300302327261e-05,
"loss": 1.285,
"step": 14580
},
{
"epoch": 0.7,
"grad_norm": 0.3847843110561371,
"learning_rate": 3.636621804291594e-05,
"loss": 1.2971,
"step": 14590
},
{
"epoch": 0.7,
"grad_norm": 0.3598523437976837,
"learning_rate": 3.634942660140075e-05,
"loss": 1.2611,
"step": 14600
},
{
"epoch": 0.7,
"grad_norm": 0.3886086344718933,
"learning_rate": 3.633262870827217e-05,
"loss": 1.3496,
"step": 14610
},
{
"epoch": 0.7,
"grad_norm": 0.38230636715888977,
"learning_rate": 3.6315824373079003e-05,
"loss": 1.4147,
"step": 14620
},
{
"epoch": 0.7,
"grad_norm": 0.39894354343414307,
"learning_rate": 3.6299013605373744e-05,
"loss": 1.2805,
"step": 14630
},
{
"epoch": 0.7,
"grad_norm": 0.36072587966918945,
"learning_rate": 3.628219641471249e-05,
"loss": 1.3559,
"step": 14640
},
{
"epoch": 0.7,
"grad_norm": 0.37694451212882996,
"learning_rate": 3.6265372810655054e-05,
"loss": 1.3956,
"step": 14650
},
{
"epoch": 0.7,
"grad_norm": 0.4612378180027008,
"learning_rate": 3.624854280276484e-05,
"loss": 1.3273,
"step": 14660
},
{
"epoch": 0.7,
"grad_norm": 0.4327034652233124,
"learning_rate": 3.623170640060891e-05,
"loss": 1.2794,
"step": 14670
},
{
"epoch": 0.7,
"grad_norm": 0.3651736378669739,
"learning_rate": 3.621486361375799e-05,
"loss": 1.3916,
"step": 14680
},
{
"epoch": 0.71,
"grad_norm": 0.34631049633026123,
"learning_rate": 3.619801445178637e-05,
"loss": 1.2983,
"step": 14690
},
{
"epoch": 0.71,
"grad_norm": 0.37691110372543335,
"learning_rate": 3.6181158924272035e-05,
"loss": 1.2857,
"step": 14700
},
{
"epoch": 0.71,
"grad_norm": 0.42540687322616577,
"learning_rate": 3.616429704079654e-05,
"loss": 1.313,
"step": 14710
},
{
"epoch": 0.71,
"grad_norm": 0.405963659286499,
"learning_rate": 3.6147428810945084e-05,
"loss": 1.2947,
"step": 14720
},
{
"epoch": 0.71,
"grad_norm": 0.44601908326148987,
"learning_rate": 3.613055424430645e-05,
"loss": 1.2562,
"step": 14730
},
{
"epoch": 0.71,
"grad_norm": 0.3817850649356842,
"learning_rate": 3.611367335047302e-05,
"loss": 1.3103,
"step": 14740
},
{
"epoch": 0.71,
"grad_norm": 0.3699060082435608,
"learning_rate": 3.6096786139040815e-05,
"loss": 1.296,
"step": 14750
},
{
"epoch": 0.71,
"grad_norm": 0.32550427317619324,
"learning_rate": 3.60798926196094e-05,
"loss": 1.3523,
"step": 14760
},
{
"epoch": 0.71,
"grad_norm": 0.455151230096817,
"learning_rate": 3.606299280178195e-05,
"loss": 1.2448,
"step": 14770
},
{
"epoch": 0.71,
"grad_norm": 0.43262967467308044,
"learning_rate": 3.604608669516521e-05,
"loss": 1.3944,
"step": 14780
},
{
"epoch": 0.71,
"grad_norm": 0.38703733682632446,
"learning_rate": 3.602917430936949e-05,
"loss": 1.3769,
"step": 14790
},
{
"epoch": 0.71,
"grad_norm": 0.33776235580444336,
"learning_rate": 3.601225565400871e-05,
"loss": 1.3011,
"step": 14800
},
{
"epoch": 0.71,
"grad_norm": 0.40187305212020874,
"learning_rate": 3.599533073870031e-05,
"loss": 1.2711,
"step": 14810
},
{
"epoch": 0.71,
"grad_norm": 0.37562501430511475,
"learning_rate": 3.5978399573065316e-05,
"loss": 1.3683,
"step": 14820
},
{
"epoch": 0.71,
"grad_norm": 0.336889386177063,
"learning_rate": 3.5961462166728285e-05,
"loss": 1.3193,
"step": 14830
},
{
"epoch": 0.71,
"grad_norm": 0.3487083911895752,
"learning_rate": 3.594451852931733e-05,
"loss": 1.3411,
"step": 14840
},
{
"epoch": 0.71,
"grad_norm": 0.41965726017951965,
"learning_rate": 3.592756867046413e-05,
"loss": 1.3481,
"step": 14850
},
{
"epoch": 0.71,
"grad_norm": 0.5000163912773132,
"learning_rate": 3.591061259980386e-05,
"loss": 1.4069,
"step": 14860
},
{
"epoch": 0.71,
"grad_norm": 0.3983001708984375,
"learning_rate": 3.589365032697526e-05,
"loss": 1.3874,
"step": 14870
},
{
"epoch": 0.71,
"grad_norm": 0.48774582147598267,
"learning_rate": 3.587668186162057e-05,
"loss": 1.342,
"step": 14880
},
{
"epoch": 0.71,
"grad_norm": 0.3706805408000946,
"learning_rate": 3.585970721338556e-05,
"loss": 1.2916,
"step": 14890
},
{
"epoch": 0.72,
"grad_norm": 0.33413049578666687,
"learning_rate": 3.5842726391919535e-05,
"loss": 1.361,
"step": 14900
},
{
"epoch": 0.72,
"grad_norm": 0.4676301181316376,
"learning_rate": 3.5825739406875276e-05,
"loss": 1.234,
"step": 14910
},
{
"epoch": 0.72,
"grad_norm": 0.4327121078968048,
"learning_rate": 3.5808746267909075e-05,
"loss": 1.4036,
"step": 14920
},
{
"epoch": 0.72,
"grad_norm": 0.3514186143875122,
"learning_rate": 3.579174698468075e-05,
"loss": 1.2973,
"step": 14930
},
{
"epoch": 0.72,
"grad_norm": 0.340087354183197,
"learning_rate": 3.577474156685358e-05,
"loss": 1.3704,
"step": 14940
},
{
"epoch": 0.72,
"grad_norm": 0.4281562268733978,
"learning_rate": 3.575773002409434e-05,
"loss": 1.2848,
"step": 14950
},
{
"epoch": 0.72,
"grad_norm": 0.3562680184841156,
"learning_rate": 3.5740712366073295e-05,
"loss": 1.3457,
"step": 14960
},
{
"epoch": 0.72,
"grad_norm": 0.412727952003479,
"learning_rate": 3.572368860246417e-05,
"loss": 1.3773,
"step": 14970
},
{
"epoch": 0.72,
"grad_norm": 0.31067582964897156,
"learning_rate": 3.5706658742944185e-05,
"loss": 1.3733,
"step": 14980
},
{
"epoch": 0.72,
"grad_norm": 0.4103313386440277,
"learning_rate": 3.5689622797194e-05,
"loss": 1.3429,
"step": 14990
},
{
"epoch": 0.72,
"grad_norm": 0.4147751033306122,
"learning_rate": 3.5672580774897765e-05,
"loss": 1.3794,
"step": 15000
},
{
"epoch": 0.72,
"grad_norm": 0.44084811210632324,
"learning_rate": 3.565553268574304e-05,
"loss": 1.3966,
"step": 15010
},
{
"epoch": 0.72,
"grad_norm": 0.352255642414093,
"learning_rate": 3.563847853942088e-05,
"loss": 1.2884,
"step": 15020
},
{
"epoch": 0.72,
"grad_norm": 0.38386282324790955,
"learning_rate": 3.562141834562574e-05,
"loss": 1.3459,
"step": 15030
},
{
"epoch": 0.72,
"grad_norm": 0.3103056252002716,
"learning_rate": 3.560435211405555e-05,
"loss": 1.3377,
"step": 15040
},
{
"epoch": 0.72,
"grad_norm": 0.4245316684246063,
"learning_rate": 3.558727985441167e-05,
"loss": 1.414,
"step": 15050
},
{
"epoch": 0.72,
"grad_norm": 0.4189935326576233,
"learning_rate": 3.557020157639884e-05,
"loss": 1.3581,
"step": 15060
},
{
"epoch": 0.72,
"grad_norm": 0.33732885122299194,
"learning_rate": 3.555311728972529e-05,
"loss": 1.396,
"step": 15070
},
{
"epoch": 0.72,
"grad_norm": 0.34590816497802734,
"learning_rate": 3.5536027004102615e-05,
"loss": 1.3512,
"step": 15080
},
{
"epoch": 0.72,
"grad_norm": 0.37284255027770996,
"learning_rate": 3.551893072924582e-05,
"loss": 1.3406,
"step": 15090
},
{
"epoch": 0.72,
"grad_norm": 0.3281930685043335,
"learning_rate": 3.550182847487336e-05,
"loss": 1.2945,
"step": 15100
},
{
"epoch": 0.73,
"grad_norm": 0.34245938062667847,
"learning_rate": 3.5484720250707045e-05,
"loss": 1.3039,
"step": 15110
},
{
"epoch": 0.73,
"grad_norm": 0.5333119034767151,
"learning_rate": 3.546760606647209e-05,
"loss": 1.3532,
"step": 15120
},
{
"epoch": 0.73,
"grad_norm": 0.4237481355667114,
"learning_rate": 3.5450485931897105e-05,
"loss": 1.3613,
"step": 15130
},
{
"epoch": 0.73,
"grad_norm": 0.386153906583786,
"learning_rate": 3.543335985671408e-05,
"loss": 1.3109,
"step": 15140
},
{
"epoch": 0.73,
"grad_norm": 0.41039368510246277,
"learning_rate": 3.541622785065837e-05,
"loss": 1.3117,
"step": 15150
},
{
"epoch": 0.73,
"grad_norm": 0.2796619236469269,
"learning_rate": 3.5399089923468734e-05,
"loss": 1.3218,
"step": 15160
},
{
"epoch": 0.73,
"grad_norm": 0.41052407026290894,
"learning_rate": 3.538194608488726e-05,
"loss": 1.2787,
"step": 15170
},
{
"epoch": 0.73,
"grad_norm": 0.3037787973880768,
"learning_rate": 3.536479634465941e-05,
"loss": 1.2993,
"step": 15180
},
{
"epoch": 0.73,
"grad_norm": 0.33063992857933044,
"learning_rate": 3.534764071253399e-05,
"loss": 1.3925,
"step": 15190
},
{
"epoch": 0.73,
"grad_norm": 0.3610629737377167,
"learning_rate": 3.53304791982632e-05,
"loss": 1.3809,
"step": 15200
},
{
"epoch": 0.73,
"grad_norm": 0.3680621385574341,
"learning_rate": 3.5313311811602524e-05,
"loss": 1.3217,
"step": 15210
},
{
"epoch": 0.73,
"grad_norm": 0.41854211688041687,
"learning_rate": 3.529613856231083e-05,
"loss": 1.2926,
"step": 15220
},
{
"epoch": 0.73,
"grad_norm": 0.41760143637657166,
"learning_rate": 3.5278959460150276e-05,
"loss": 1.3224,
"step": 15230
},
{
"epoch": 0.73,
"grad_norm": 0.32895025610923767,
"learning_rate": 3.526177451488639e-05,
"loss": 1.2713,
"step": 15240
},
{
"epoch": 0.73,
"grad_norm": 0.34725242853164673,
"learning_rate": 3.5244583736288e-05,
"loss": 1.2839,
"step": 15250
},
{
"epoch": 0.73,
"grad_norm": 0.42915984988212585,
"learning_rate": 3.5227387134127246e-05,
"loss": 1.352,
"step": 15260
},
{
"epoch": 0.73,
"grad_norm": 0.3569702208042145,
"learning_rate": 3.521018471817959e-05,
"loss": 1.3413,
"step": 15270
},
{
"epoch": 0.73,
"grad_norm": 0.34027430415153503,
"learning_rate": 3.519297649822379e-05,
"loss": 1.3775,
"step": 15280
},
{
"epoch": 0.73,
"grad_norm": 0.30371078848838806,
"learning_rate": 3.51757624840419e-05,
"loss": 1.2522,
"step": 15290
},
{
"epoch": 0.73,
"grad_norm": 0.316923052072525,
"learning_rate": 3.515854268541929e-05,
"loss": 1.3776,
"step": 15300
},
{
"epoch": 0.73,
"grad_norm": 0.38221609592437744,
"learning_rate": 3.51413171121446e-05,
"loss": 1.4202,
"step": 15310
},
{
"epoch": 0.74,
"grad_norm": 0.3114246726036072,
"learning_rate": 3.512408577400975e-05,
"loss": 1.3371,
"step": 15320
},
{
"epoch": 0.74,
"grad_norm": 0.3160313665866852,
"learning_rate": 3.510684868080994e-05,
"loss": 1.3509,
"step": 15330
},
{
"epoch": 0.74,
"grad_norm": 0.27797931432724,
"learning_rate": 3.508960584234364e-05,
"loss": 1.3243,
"step": 15340
},
{
"epoch": 0.74,
"grad_norm": 0.35110726952552795,
"learning_rate": 3.507235726841261e-05,
"loss": 1.2711,
"step": 15350
},
{
"epoch": 0.74,
"grad_norm": 0.40011414885520935,
"learning_rate": 3.5055102968821826e-05,
"loss": 1.3244,
"step": 15360
},
{
"epoch": 0.74,
"grad_norm": 0.3373492360115051,
"learning_rate": 3.5037842953379574e-05,
"loss": 1.2898,
"step": 15370
},
{
"epoch": 0.74,
"grad_norm": 0.3832080066204071,
"learning_rate": 3.5020577231897325e-05,
"loss": 1.2819,
"step": 15380
},
{
"epoch": 0.74,
"grad_norm": 0.34173035621643066,
"learning_rate": 3.500330581418984e-05,
"loss": 1.3832,
"step": 15390
},
{
"epoch": 0.74,
"grad_norm": 0.35764646530151367,
"learning_rate": 3.498602871007511e-05,
"loss": 1.306,
"step": 15400
},
{
"epoch": 0.74,
"grad_norm": 0.41290393471717834,
"learning_rate": 3.496874592937434e-05,
"loss": 1.3033,
"step": 15410
},
{
"epoch": 0.74,
"grad_norm": 0.3148930072784424,
"learning_rate": 3.4951457481912e-05,
"loss": 1.4381,
"step": 15420
},
{
"epoch": 0.74,
"grad_norm": 0.35221928358078003,
"learning_rate": 3.493416337751573e-05,
"loss": 1.3796,
"step": 15430
},
{
"epoch": 0.74,
"grad_norm": 0.392743855714798,
"learning_rate": 3.491686362601642e-05,
"loss": 1.3034,
"step": 15440
},
{
"epoch": 0.74,
"grad_norm": 0.47063612937927246,
"learning_rate": 3.490128902952176e-05,
"loss": 1.3629,
"step": 15450
},
{
"epoch": 0.74,
"grad_norm": 0.3460381329059601,
"learning_rate": 3.488397857562226e-05,
"loss": 1.3236,
"step": 15460
},
{
"epoch": 0.74,
"grad_norm": 0.5289183259010315,
"learning_rate": 3.48666625031474e-05,
"loss": 1.3391,
"step": 15470
},
{
"epoch": 0.74,
"grad_norm": 0.34871774911880493,
"learning_rate": 3.484934082194058e-05,
"loss": 1.3132,
"step": 15480
},
{
"epoch": 0.74,
"grad_norm": 0.32303252816200256,
"learning_rate": 3.4832013541848335e-05,
"loss": 1.3117,
"step": 15490
},
{
"epoch": 0.74,
"grad_norm": 0.38771501183509827,
"learning_rate": 3.481468067272041e-05,
"loss": 1.41,
"step": 15500
},
{
"epoch": 0.74,
"grad_norm": 0.4492136538028717,
"learning_rate": 3.479734222440973e-05,
"loss": 1.3648,
"step": 15510
},
{
"epoch": 0.74,
"grad_norm": 0.506844699382782,
"learning_rate": 3.4781732858874854e-05,
"loss": 1.389,
"step": 15520
},
{
"epoch": 0.75,
"grad_norm": 0.3654049336910248,
"learning_rate": 3.476438383727308e-05,
"loss": 1.4141,
"step": 15530
},
{
"epoch": 0.75,
"grad_norm": 0.35734307765960693,
"learning_rate": 3.474702926507992e-05,
"loss": 1.3193,
"step": 15540
},
{
"epoch": 0.75,
"grad_norm": 0.4684094488620758,
"learning_rate": 3.4729669152160635e-05,
"loss": 1.377,
"step": 15550
},
{
"epoch": 0.75,
"grad_norm": 0.5034934878349304,
"learning_rate": 3.471230350838362e-05,
"loss": 1.291,
"step": 15560
},
{
"epoch": 0.75,
"grad_norm": 0.3501163125038147,
"learning_rate": 3.469493234362044e-05,
"loss": 1.2315,
"step": 15570
},
{
"epoch": 0.75,
"grad_norm": 0.33607709407806396,
"learning_rate": 3.4677555667745785e-05,
"loss": 1.3758,
"step": 15580
},
{
"epoch": 0.75,
"grad_norm": 0.42021626234054565,
"learning_rate": 3.4660173490637455e-05,
"loss": 1.3094,
"step": 15590
},
{
"epoch": 0.75,
"grad_norm": 0.37869101762771606,
"learning_rate": 3.4642785822176405e-05,
"loss": 1.3697,
"step": 15600
},
{
"epoch": 0.75,
"grad_norm": 0.45084768533706665,
"learning_rate": 3.462539267224671e-05,
"loss": 1.2666,
"step": 15610
},
{
"epoch": 0.75,
"grad_norm": 0.2922057509422302,
"learning_rate": 3.460799405073555e-05,
"loss": 1.3234,
"step": 15620
},
{
"epoch": 0.75,
"grad_norm": 0.3410915732383728,
"learning_rate": 3.459058996753322e-05,
"loss": 1.3302,
"step": 15630
},
{
"epoch": 0.75,
"grad_norm": 0.46749573945999146,
"learning_rate": 3.457318043253313e-05,
"loss": 1.2703,
"step": 15640
},
{
"epoch": 0.75,
"grad_norm": 0.3407883644104004,
"learning_rate": 3.455576545563176e-05,
"loss": 1.3228,
"step": 15650
},
{
"epoch": 0.75,
"grad_norm": 0.4112377166748047,
"learning_rate": 3.453834504672872e-05,
"loss": 1.3712,
"step": 15660
},
{
"epoch": 0.75,
"grad_norm": 0.40037932991981506,
"learning_rate": 3.452091921572667e-05,
"loss": 1.2753,
"step": 15670
},
{
"epoch": 0.75,
"grad_norm": 0.4184977412223816,
"learning_rate": 3.450348797253139e-05,
"loss": 1.3413,
"step": 15680
},
{
"epoch": 0.75,
"grad_norm": 0.4081215262413025,
"learning_rate": 3.448605132705173e-05,
"loss": 1.2991,
"step": 15690
},
{
"epoch": 0.75,
"grad_norm": 0.37343844771385193,
"learning_rate": 3.4468609289199574e-05,
"loss": 1.3243,
"step": 15700
},
{
"epoch": 0.75,
"grad_norm": 0.3959372639656067,
"learning_rate": 3.44511618688899e-05,
"loss": 1.3731,
"step": 15710
},
{
"epoch": 0.75,
"grad_norm": 0.3384896218776703,
"learning_rate": 3.443370907604076e-05,
"loss": 1.3901,
"step": 15720
},
{
"epoch": 0.76,
"grad_norm": 0.3795836269855499,
"learning_rate": 3.4416250920573236e-05,
"loss": 1.3871,
"step": 15730
},
{
"epoch": 0.76,
"grad_norm": 0.3807082772254944,
"learning_rate": 3.439878741241146e-05,
"loss": 1.3264,
"step": 15740
},
{
"epoch": 0.76,
"grad_norm": 0.319669634103775,
"learning_rate": 3.4381318561482614e-05,
"loss": 1.3729,
"step": 15750
},
{
"epoch": 0.76,
"grad_norm": 0.39579904079437256,
"learning_rate": 3.4363844377716926e-05,
"loss": 1.3691,
"step": 15760
},
{
"epoch": 0.76,
"grad_norm": 0.347087025642395,
"learning_rate": 3.434636487104763e-05,
"loss": 1.302,
"step": 15770
},
{
"epoch": 0.76,
"grad_norm": 0.38257354497909546,
"learning_rate": 3.4328880051411e-05,
"loss": 1.4187,
"step": 15780
},
{
"epoch": 0.76,
"grad_norm": 0.33336159586906433,
"learning_rate": 3.4311389928746346e-05,
"loss": 1.2813,
"step": 15790
},
{
"epoch": 0.76,
"grad_norm": 0.48640888929367065,
"learning_rate": 3.429389451299596e-05,
"loss": 1.341,
"step": 15800
},
{
"epoch": 0.76,
"grad_norm": 0.3425988256931305,
"learning_rate": 3.427639381410518e-05,
"loss": 1.3247,
"step": 15810
},
{
"epoch": 0.76,
"grad_norm": 0.3972455859184265,
"learning_rate": 3.42588878420223e-05,
"loss": 1.3641,
"step": 15820
},
{
"epoch": 0.76,
"grad_norm": 0.3746291995048523,
"learning_rate": 3.424137660669867e-05,
"loss": 1.3359,
"step": 15830
},
{
"epoch": 0.76,
"grad_norm": 0.34066423773765564,
"learning_rate": 3.4223860118088577e-05,
"loss": 1.279,
"step": 15840
},
{
"epoch": 0.76,
"grad_norm": 0.33098939061164856,
"learning_rate": 3.420633838614933e-05,
"loss": 1.307,
"step": 15850
},
{
"epoch": 0.76,
"grad_norm": 0.4499354958534241,
"learning_rate": 3.41888114208412e-05,
"loss": 1.3243,
"step": 15860
},
{
"epoch": 0.76,
"grad_norm": 0.320658415555954,
"learning_rate": 3.417127923212745e-05,
"loss": 1.2754,
"step": 15870
},
{
"epoch": 0.76,
"grad_norm": 0.3453558385372162,
"learning_rate": 3.41537418299743e-05,
"loss": 1.3245,
"step": 15880
},
{
"epoch": 0.76,
"grad_norm": 0.3926355242729187,
"learning_rate": 3.413619922435092e-05,
"loss": 1.2863,
"step": 15890
},
{
"epoch": 0.76,
"grad_norm": 0.28055837750434875,
"learning_rate": 3.411865142522947e-05,
"loss": 1.4284,
"step": 15900
},
{
"epoch": 0.76,
"grad_norm": 0.5162535905838013,
"learning_rate": 3.410109844258503e-05,
"loss": 1.3172,
"step": 15910
},
{
"epoch": 0.76,
"grad_norm": 0.3829997479915619,
"learning_rate": 3.408354028639565e-05,
"loss": 1.3201,
"step": 15920
},
{
"epoch": 0.76,
"grad_norm": 0.3923022747039795,
"learning_rate": 3.4065976966642324e-05,
"loss": 1.3368,
"step": 15930
},
{
"epoch": 0.77,
"grad_norm": 0.3313416838645935,
"learning_rate": 3.4048408493308966e-05,
"loss": 1.4597,
"step": 15940
},
{
"epoch": 0.77,
"grad_norm": 0.37384408712387085,
"learning_rate": 3.4030834876382414e-05,
"loss": 1.3024,
"step": 15950
},
{
"epoch": 0.77,
"grad_norm": 0.3159520924091339,
"learning_rate": 3.401325612585244e-05,
"loss": 1.3996,
"step": 15960
},
{
"epoch": 0.77,
"grad_norm": 0.31553277373313904,
"learning_rate": 3.399567225171174e-05,
"loss": 1.2986,
"step": 15970
},
{
"epoch": 0.77,
"grad_norm": 0.37068814039230347,
"learning_rate": 3.397808326395591e-05,
"loss": 1.3183,
"step": 15980
},
{
"epoch": 0.77,
"grad_norm": 0.34177079796791077,
"learning_rate": 3.396048917258348e-05,
"loss": 1.3074,
"step": 15990
},
{
"epoch": 0.77,
"grad_norm": 0.32131263613700867,
"learning_rate": 3.3942889987595835e-05,
"loss": 1.3606,
"step": 16000
},
{
"epoch": 0.77,
"grad_norm": 0.4284037947654724,
"learning_rate": 3.3925285718997294e-05,
"loss": 1.2924,
"step": 16010
},
{
"epoch": 0.77,
"grad_norm": 0.3265150487422943,
"learning_rate": 3.390767637679505e-05,
"loss": 1.3406,
"step": 16020
},
{
"epoch": 0.77,
"grad_norm": 0.43392255902290344,
"learning_rate": 3.389006197099919e-05,
"loss": 1.3588,
"step": 16030
},
{
"epoch": 0.77,
"grad_norm": 0.31104499101638794,
"learning_rate": 3.387244251162266e-05,
"loss": 1.2772,
"step": 16040
},
{
"epoch": 0.77,
"grad_norm": 0.30960074067115784,
"learning_rate": 3.385481800868129e-05,
"loss": 1.306,
"step": 16050
},
{
"epoch": 0.77,
"grad_norm": 0.3950722813606262,
"learning_rate": 3.38371884721938e-05,
"loss": 1.3437,
"step": 16060
},
{
"epoch": 0.77,
"grad_norm": 0.32086747884750366,
"learning_rate": 3.381955391218172e-05,
"loss": 1.2928,
"step": 16070
},
{
"epoch": 0.77,
"grad_norm": 0.3770638108253479,
"learning_rate": 3.3801914338669486e-05,
"loss": 1.3603,
"step": 16080
},
{
"epoch": 0.77,
"grad_norm": 0.352895051240921,
"learning_rate": 3.378426976168436e-05,
"loss": 1.31,
"step": 16090
},
{
"epoch": 0.77,
"grad_norm": 0.41496115922927856,
"learning_rate": 3.3766620191256446e-05,
"loss": 1.2881,
"step": 16100
},
{
"epoch": 0.77,
"grad_norm": 0.41254112124443054,
"learning_rate": 3.3748965637418696e-05,
"loss": 1.3398,
"step": 16110
},
{
"epoch": 0.77,
"grad_norm": 0.3925231099128723,
"learning_rate": 3.3731306110206884e-05,
"loss": 1.2106,
"step": 16120
},
{
"epoch": 0.77,
"grad_norm": 0.40501669049263,
"learning_rate": 3.371364161965963e-05,
"loss": 1.244,
"step": 16130
},
{
"epoch": 0.77,
"grad_norm": 0.3687911927700043,
"learning_rate": 3.3695972175818345e-05,
"loss": 1.3528,
"step": 16140
},
{
"epoch": 0.78,
"grad_norm": 0.4450591802597046,
"learning_rate": 3.367829778872729e-05,
"loss": 1.3228,
"step": 16150
},
{
"epoch": 0.78,
"grad_norm": 0.41986939311027527,
"learning_rate": 3.3660618468433516e-05,
"loss": 1.3023,
"step": 16160
},
{
"epoch": 0.78,
"grad_norm": 0.3356514573097229,
"learning_rate": 3.3642934224986875e-05,
"loss": 1.2859,
"step": 16170
},
{
"epoch": 0.78,
"grad_norm": 0.45237648487091064,
"learning_rate": 3.362524506844003e-05,
"loss": 1.3536,
"step": 16180
},
{
"epoch": 0.78,
"grad_norm": 0.38273078203201294,
"learning_rate": 3.360755100884844e-05,
"loss": 1.2678,
"step": 16190
},
{
"epoch": 0.78,
"grad_norm": 0.45566296577453613,
"learning_rate": 3.358985205627032e-05,
"loss": 1.3944,
"step": 16200
},
{
"epoch": 0.78,
"grad_norm": 0.38295426964759827,
"learning_rate": 3.357214822076672e-05,
"loss": 1.2928,
"step": 16210
},
{
"epoch": 0.78,
"grad_norm": 0.35819345712661743,
"learning_rate": 3.3554439512401406e-05,
"loss": 1.3018,
"step": 16220
},
{
"epoch": 0.78,
"grad_norm": 0.3953298330307007,
"learning_rate": 3.353672594124096e-05,
"loss": 1.358,
"step": 16230
},
{
"epoch": 0.78,
"grad_norm": 0.4462447762489319,
"learning_rate": 3.3519007517354715e-05,
"loss": 1.3645,
"step": 16240
},
{
"epoch": 0.78,
"grad_norm": 0.36345192790031433,
"learning_rate": 3.350128425081474e-05,
"loss": 1.2657,
"step": 16250
},
{
"epoch": 0.78,
"grad_norm": 0.44041749835014343,
"learning_rate": 3.3483556151695895e-05,
"loss": 1.2849,
"step": 16260
},
{
"epoch": 0.78,
"grad_norm": 0.3163689970970154,
"learning_rate": 3.3465823230075755e-05,
"loss": 1.3429,
"step": 16270
},
{
"epoch": 0.78,
"grad_norm": 0.4109678268432617,
"learning_rate": 3.344808549603466e-05,
"loss": 1.3326,
"step": 16280
},
{
"epoch": 0.78,
"grad_norm": 0.41100144386291504,
"learning_rate": 3.343034295965566e-05,
"loss": 1.3092,
"step": 16290
},
{
"epoch": 0.78,
"grad_norm": 0.43808257579803467,
"learning_rate": 3.3412595631024555e-05,
"loss": 1.3103,
"step": 16300
},
{
"epoch": 0.78,
"grad_norm": 0.41964903473854065,
"learning_rate": 3.339484352022987e-05,
"loss": 1.2804,
"step": 16310
},
{
"epoch": 0.78,
"grad_norm": 0.602850615978241,
"learning_rate": 3.337708663736284e-05,
"loss": 1.3841,
"step": 16320
},
{
"epoch": 0.78,
"grad_norm": 0.3914157748222351,
"learning_rate": 3.335932499251742e-05,
"loss": 1.2953,
"step": 16330
},
{
"epoch": 0.78,
"grad_norm": 0.29213619232177734,
"learning_rate": 3.334155859579025e-05,
"loss": 1.2834,
"step": 16340
},
{
"epoch": 0.78,
"grad_norm": 0.4115000069141388,
"learning_rate": 3.332378745728069e-05,
"loss": 1.2945,
"step": 16350
},
{
"epoch": 0.79,
"grad_norm": 0.3223179876804352,
"learning_rate": 3.3306011587090814e-05,
"loss": 1.3647,
"step": 16360
},
{
"epoch": 0.79,
"grad_norm": 0.34926676750183105,
"learning_rate": 3.328823099532535e-05,
"loss": 1.3465,
"step": 16370
},
{
"epoch": 0.79,
"grad_norm": 0.37329503893852234,
"learning_rate": 3.327044569209173e-05,
"loss": 1.4502,
"step": 16380
},
{
"epoch": 0.79,
"grad_norm": 0.340608686208725,
"learning_rate": 3.325443489923214e-05,
"loss": 1.4363,
"step": 16390
},
{
"epoch": 0.79,
"grad_norm": 0.4019469916820526,
"learning_rate": 3.3236640672064566e-05,
"loss": 1.374,
"step": 16400
},
{
"epoch": 0.79,
"grad_norm": 0.31388214230537415,
"learning_rate": 3.321884176275551e-05,
"loss": 1.3856,
"step": 16410
},
{
"epoch": 0.79,
"grad_norm": 0.4253236651420593,
"learning_rate": 3.320103818142278e-05,
"loss": 1.3258,
"step": 16420
},
{
"epoch": 0.79,
"grad_norm": 0.4024079442024231,
"learning_rate": 3.3183229938186896e-05,
"loss": 1.2828,
"step": 16430
},
{
"epoch": 0.79,
"grad_norm": 0.33157557249069214,
"learning_rate": 3.316541704317102e-05,
"loss": 1.3197,
"step": 16440
},
{
"epoch": 0.79,
"grad_norm": 0.30166301131248474,
"learning_rate": 3.3147599506500906e-05,
"loss": 1.3045,
"step": 16450
},
{
"epoch": 0.79,
"grad_norm": 0.3619363307952881,
"learning_rate": 3.312977733830499e-05,
"loss": 1.3139,
"step": 16460
},
{
"epoch": 0.79,
"grad_norm": 0.42440348863601685,
"learning_rate": 3.3111950548714354e-05,
"loss": 1.3376,
"step": 16470
},
{
"epoch": 0.79,
"grad_norm": 0.43252623081207275,
"learning_rate": 3.3094119147862677e-05,
"loss": 1.3633,
"step": 16480
},
{
"epoch": 0.79,
"grad_norm": 0.3545466959476471,
"learning_rate": 3.307628314588626e-05,
"loss": 1.4885,
"step": 16490
},
{
"epoch": 0.79,
"grad_norm": 0.4845661520957947,
"learning_rate": 3.3058442552924036e-05,
"loss": 1.2848,
"step": 16500
},
{
"epoch": 0.79,
"grad_norm": 0.5006143450737,
"learning_rate": 3.3040597379117536e-05,
"loss": 1.2833,
"step": 16510
},
{
"epoch": 0.79,
"grad_norm": 0.330059289932251,
"learning_rate": 3.302274763461091e-05,
"loss": 1.4004,
"step": 16520
},
{
"epoch": 0.79,
"grad_norm": 0.36441609263420105,
"learning_rate": 3.3004893329550874e-05,
"loss": 1.3448,
"step": 16530
},
{
"epoch": 0.79,
"grad_norm": 0.3933323621749878,
"learning_rate": 3.298703447408679e-05,
"loss": 1.3773,
"step": 16540
},
{
"epoch": 0.79,
"grad_norm": 0.42539799213409424,
"learning_rate": 3.2969171078370534e-05,
"loss": 1.3145,
"step": 16550
},
{
"epoch": 0.79,
"grad_norm": 0.3406704366207123,
"learning_rate": 3.2951303152556637e-05,
"loss": 1.314,
"step": 16560
},
{
"epoch": 0.8,
"grad_norm": 0.586881160736084,
"learning_rate": 3.293343070680216e-05,
"loss": 1.2893,
"step": 16570
},
{
"epoch": 0.8,
"grad_norm": 0.3537682890892029,
"learning_rate": 3.291555375126674e-05,
"loss": 1.3279,
"step": 16580
},
{
"epoch": 0.8,
"grad_norm": 0.4622391164302826,
"learning_rate": 3.289767229611259e-05,
"loss": 1.3439,
"step": 16590
},
{
"epoch": 0.8,
"grad_norm": 0.40260565280914307,
"learning_rate": 3.287978635150447e-05,
"loss": 1.3218,
"step": 16600
},
{
"epoch": 0.8,
"grad_norm": 0.3612897992134094,
"learning_rate": 3.286189592760969e-05,
"loss": 1.3182,
"step": 16610
},
{
"epoch": 0.8,
"grad_norm": 0.3406197726726532,
"learning_rate": 3.284400103459812e-05,
"loss": 1.2708,
"step": 16620
},
{
"epoch": 0.8,
"grad_norm": 0.3955172002315521,
"learning_rate": 3.2826101682642165e-05,
"loss": 1.3081,
"step": 16630
},
{
"epoch": 0.8,
"grad_norm": 0.42072418332099915,
"learning_rate": 3.280819788191675e-05,
"loss": 1.2815,
"step": 16640
},
{
"epoch": 0.8,
"grad_norm": 0.41923463344573975,
"learning_rate": 3.279028964259935e-05,
"loss": 1.3194,
"step": 16650
},
{
"epoch": 0.8,
"grad_norm": 0.58099365234375,
"learning_rate": 3.277237697486995e-05,
"loss": 1.3009,
"step": 16660
},
{
"epoch": 0.8,
"grad_norm": 0.5021660327911377,
"learning_rate": 3.2754459888911057e-05,
"loss": 1.3691,
"step": 16670
},
{
"epoch": 0.8,
"grad_norm": 0.36260437965393066,
"learning_rate": 3.273653839490769e-05,
"loss": 1.3105,
"step": 16680
},
{
"epoch": 0.8,
"grad_norm": 0.42414990067481995,
"learning_rate": 3.2718612503047373e-05,
"loss": 1.316,
"step": 16690
},
{
"epoch": 0.8,
"grad_norm": 0.3988606929779053,
"learning_rate": 3.2700682223520124e-05,
"loss": 1.3713,
"step": 16700
},
{
"epoch": 0.8,
"grad_norm": 0.36285969614982605,
"learning_rate": 3.2682747566518464e-05,
"loss": 1.3237,
"step": 16710
},
{
"epoch": 0.8,
"grad_norm": 0.30997583270072937,
"learning_rate": 3.26648085422374e-05,
"loss": 1.3455,
"step": 16720
},
{
"epoch": 0.8,
"grad_norm": 0.5143419504165649,
"learning_rate": 3.264686516087442e-05,
"loss": 1.2628,
"step": 16730
},
{
"epoch": 0.8,
"grad_norm": 0.3977755904197693,
"learning_rate": 3.26289174326295e-05,
"loss": 1.312,
"step": 16740
},
{
"epoch": 0.8,
"grad_norm": 0.45780009031295776,
"learning_rate": 3.2610965367705054e-05,
"loss": 1.2683,
"step": 16750
},
{
"epoch": 0.8,
"grad_norm": 0.32445719838142395,
"learning_rate": 3.2593008976306e-05,
"loss": 1.3102,
"step": 16760
},
{
"epoch": 0.8,
"grad_norm": 0.44369634985923767,
"learning_rate": 3.257504826863969e-05,
"loss": 1.3707,
"step": 16770
},
{
"epoch": 0.81,
"grad_norm": 0.4420567750930786,
"learning_rate": 3.255708325491595e-05,
"loss": 1.2317,
"step": 16780
},
{
"epoch": 0.81,
"grad_norm": 0.500679075717926,
"learning_rate": 3.2539113945347035e-05,
"loss": 1.3552,
"step": 16790
},
{
"epoch": 0.81,
"grad_norm": 0.36973854899406433,
"learning_rate": 3.252114035014765e-05,
"loss": 1.3212,
"step": 16800
},
{
"epoch": 0.81,
"grad_norm": 0.27142518758773804,
"learning_rate": 3.250316247953493e-05,
"loss": 1.4584,
"step": 16810
},
{
"epoch": 0.81,
"grad_norm": 0.39117828011512756,
"learning_rate": 3.248518034372846e-05,
"loss": 1.3993,
"step": 16820
},
{
"epoch": 0.81,
"grad_norm": 0.3525172770023346,
"learning_rate": 3.2467193952950223e-05,
"loss": 1.4157,
"step": 16830
},
{
"epoch": 0.81,
"grad_norm": 0.36123204231262207,
"learning_rate": 3.244920331742465e-05,
"loss": 1.3388,
"step": 16840
},
{
"epoch": 0.81,
"grad_norm": 0.46618422865867615,
"learning_rate": 3.243120844737854e-05,
"loss": 1.3456,
"step": 16850
},
{
"epoch": 0.81,
"grad_norm": 0.3534936308860779,
"learning_rate": 3.241320935304115e-05,
"loss": 1.316,
"step": 16860
},
{
"epoch": 0.81,
"grad_norm": 0.40895065665245056,
"learning_rate": 3.239520604464411e-05,
"loss": 1.3641,
"step": 16870
},
{
"epoch": 0.81,
"grad_norm": 0.37877899408340454,
"learning_rate": 3.237719853242145e-05,
"loss": 1.337,
"step": 16880
},
{
"epoch": 0.81,
"grad_norm": 0.39393922686576843,
"learning_rate": 3.235918682660959e-05,
"loss": 1.3191,
"step": 16890
},
{
"epoch": 0.81,
"grad_norm": 0.32107535004615784,
"learning_rate": 3.234117093744734e-05,
"loss": 1.2909,
"step": 16900
},
{
"epoch": 0.81,
"grad_norm": 0.3763900101184845,
"learning_rate": 3.232315087517587e-05,
"loss": 1.3084,
"step": 16910
},
{
"epoch": 0.81,
"grad_norm": 0.4269709885120392,
"learning_rate": 3.230512665003875e-05,
"loss": 1.3903,
"step": 16920
},
{
"epoch": 0.81,
"grad_norm": 0.5338302850723267,
"learning_rate": 3.228709827228189e-05,
"loss": 1.4211,
"step": 16930
},
{
"epoch": 0.81,
"grad_norm": 0.28215596079826355,
"learning_rate": 3.2269065752153585e-05,
"loss": 1.3266,
"step": 16940
},
{
"epoch": 0.81,
"grad_norm": 0.3255442678928375,
"learning_rate": 3.225102909990445e-05,
"loss": 1.2737,
"step": 16950
},
{
"epoch": 0.81,
"grad_norm": 0.4847237467765808,
"learning_rate": 3.2232988325787494e-05,
"loss": 1.3435,
"step": 16960
},
{
"epoch": 0.81,
"grad_norm": 0.4686855673789978,
"learning_rate": 3.221494344005803e-05,
"loss": 1.3514,
"step": 16970
},
{
"epoch": 0.82,
"grad_norm": 0.3423585593700409,
"learning_rate": 3.219689445297374e-05,
"loss": 1.3089,
"step": 16980
},
{
"epoch": 0.82,
"grad_norm": 0.27913469076156616,
"learning_rate": 3.2178841374794606e-05,
"loss": 1.3397,
"step": 16990
},
{
"epoch": 0.82,
"grad_norm": 0.36152300238609314,
"learning_rate": 3.2160784215782946e-05,
"loss": 1.3537,
"step": 17000
},
{
"epoch": 0.82,
"grad_norm": 0.4597972631454468,
"learning_rate": 3.214272298620342e-05,
"loss": 1.2604,
"step": 17010
},
{
"epoch": 0.82,
"grad_norm": 0.5025787949562073,
"learning_rate": 3.212465769632298e-05,
"loss": 1.4531,
"step": 17020
},
{
"epoch": 0.82,
"grad_norm": 0.38340044021606445,
"learning_rate": 3.210658835641088e-05,
"loss": 1.4186,
"step": 17030
},
{
"epoch": 0.82,
"grad_norm": 0.491563618183136,
"learning_rate": 3.2088514976738704e-05,
"loss": 1.3074,
"step": 17040
},
{
"epoch": 0.82,
"grad_norm": 0.37494492530822754,
"learning_rate": 3.20704375675803e-05,
"loss": 1.3298,
"step": 17050
},
{
"epoch": 0.82,
"grad_norm": 0.351662814617157,
"learning_rate": 3.205235613921183e-05,
"loss": 1.3747,
"step": 17060
},
{
"epoch": 0.82,
"grad_norm": 0.3045153319835663,
"learning_rate": 3.203427070191173e-05,
"loss": 1.334,
"step": 17070
},
{
"epoch": 0.82,
"grad_norm": 0.4291684329509735,
"learning_rate": 3.2016181265960724e-05,
"loss": 1.2339,
"step": 17080
},
{
"epoch": 0.82,
"grad_norm": 0.3669149577617645,
"learning_rate": 3.199808784164179e-05,
"loss": 1.3468,
"step": 17090
},
{
"epoch": 0.82,
"grad_norm": 0.3406016230583191,
"learning_rate": 3.197999043924019e-05,
"loss": 1.3371,
"step": 17100
},
{
"epoch": 0.82,
"grad_norm": 0.3374679982662201,
"learning_rate": 3.196188906904346e-05,
"loss": 1.3117,
"step": 17110
},
{
"epoch": 0.82,
"grad_norm": 0.40612927079200745,
"learning_rate": 3.1943783741341346e-05,
"loss": 1.4017,
"step": 17120
},
{
"epoch": 0.82,
"grad_norm": 0.3344114124774933,
"learning_rate": 3.192567446642589e-05,
"loss": 1.2739,
"step": 17130
},
{
"epoch": 0.82,
"grad_norm": 0.45558154582977295,
"learning_rate": 3.1907561254591364e-05,
"loss": 1.3339,
"step": 17140
},
{
"epoch": 0.82,
"grad_norm": 0.3947155177593231,
"learning_rate": 3.188944411613426e-05,
"loss": 1.2502,
"step": 17150
},
{
"epoch": 0.82,
"grad_norm": 0.47665175795555115,
"learning_rate": 3.187132306135332e-05,
"loss": 1.2517,
"step": 17160
},
{
"epoch": 0.82,
"grad_norm": 0.3197881579399109,
"learning_rate": 3.185319810054952e-05,
"loss": 1.367,
"step": 17170
},
{
"epoch": 0.82,
"grad_norm": 0.36030516028404236,
"learning_rate": 3.183506924402603e-05,
"loss": 1.3364,
"step": 17180
},
{
"epoch": 0.83,
"grad_norm": 0.29805877804756165,
"learning_rate": 3.181693650208826e-05,
"loss": 1.3839,
"step": 17190
},
{
"epoch": 0.83,
"grad_norm": 0.33813735842704773,
"learning_rate": 3.179879988504382e-05,
"loss": 1.3912,
"step": 17200
},
{
"epoch": 0.83,
"grad_norm": 0.39367249608039856,
"learning_rate": 3.178065940320251e-05,
"loss": 1.2968,
"step": 17210
},
{
"epoch": 0.83,
"grad_norm": 0.3124770522117615,
"learning_rate": 3.176251506687633e-05,
"loss": 1.2809,
"step": 17220
},
{
"epoch": 0.83,
"grad_norm": 0.5272161960601807,
"learning_rate": 3.174436688637949e-05,
"loss": 1.3801,
"step": 17230
},
{
"epoch": 0.83,
"grad_norm": 0.33624276518821716,
"learning_rate": 3.17262148720284e-05,
"loss": 1.2705,
"step": 17240
},
{
"epoch": 0.83,
"grad_norm": 0.35133588314056396,
"learning_rate": 3.1708059034141584e-05,
"loss": 1.2909,
"step": 17250
},
{
"epoch": 0.83,
"grad_norm": 0.36546969413757324,
"learning_rate": 3.16898993830398e-05,
"loss": 1.3626,
"step": 17260
},
{
"epoch": 0.83,
"grad_norm": 0.37273362278938293,
"learning_rate": 3.167173592904593e-05,
"loss": 1.2907,
"step": 17270
},
{
"epoch": 0.83,
"grad_norm": 0.43583109974861145,
"learning_rate": 3.165356868248507e-05,
"loss": 1.3419,
"step": 17280
},
{
"epoch": 0.83,
"grad_norm": 0.537114679813385,
"learning_rate": 3.163539765368443e-05,
"loss": 1.3237,
"step": 17290
},
{
"epoch": 0.83,
"grad_norm": 0.30303865671157837,
"learning_rate": 3.1617222852973385e-05,
"loss": 1.3162,
"step": 17300
},
{
"epoch": 0.83,
"grad_norm": 0.35852643847465515,
"learning_rate": 3.1599044290683446e-05,
"loss": 1.3859,
"step": 17310
},
{
"epoch": 0.83,
"grad_norm": 0.29658326506614685,
"learning_rate": 3.1580861977148254e-05,
"loss": 1.337,
"step": 17320
},
{
"epoch": 0.83,
"grad_norm": 0.375601589679718,
"learning_rate": 3.1562675922703625e-05,
"loss": 1.3186,
"step": 17330
},
{
"epoch": 0.83,
"grad_norm": 0.38480672240257263,
"learning_rate": 3.154448613768746e-05,
"loss": 1.3463,
"step": 17340
},
{
"epoch": 0.83,
"grad_norm": 0.41698387265205383,
"learning_rate": 3.1526292632439794e-05,
"loss": 1.3551,
"step": 17350
},
{
"epoch": 0.83,
"grad_norm": 0.3713897466659546,
"learning_rate": 3.1508095417302766e-05,
"loss": 1.3709,
"step": 17360
},
{
"epoch": 0.83,
"grad_norm": 0.3444213569164276,
"learning_rate": 3.1489894502620645e-05,
"loss": 1.4015,
"step": 17370
},
{
"epoch": 0.83,
"grad_norm": 0.35246115922927856,
"learning_rate": 3.147168989873979e-05,
"loss": 1.3808,
"step": 17380
},
{
"epoch": 0.83,
"grad_norm": 0.37460118532180786,
"learning_rate": 3.145348161600867e-05,
"loss": 1.349,
"step": 17390
},
{
"epoch": 0.84,
"grad_norm": 0.3781125545501709,
"learning_rate": 3.143526966477781e-05,
"loss": 1.3862,
"step": 17400
},
{
"epoch": 0.84,
"grad_norm": 0.44478994607925415,
"learning_rate": 3.141705405539987e-05,
"loss": 1.2594,
"step": 17410
},
{
"epoch": 0.84,
"grad_norm": 0.36425256729125977,
"learning_rate": 3.139883479822954e-05,
"loss": 1.2283,
"step": 17420
},
{
"epoch": 0.84,
"grad_norm": 0.5899288058280945,
"learning_rate": 3.138061190362362e-05,
"loss": 1.3141,
"step": 17430
},
{
"epoch": 0.84,
"grad_norm": 0.36295321583747864,
"learning_rate": 3.1362385381940975e-05,
"loss": 1.2882,
"step": 17440
},
{
"epoch": 0.84,
"grad_norm": 0.5067065954208374,
"learning_rate": 3.1344155243542505e-05,
"loss": 1.3461,
"step": 17450
},
{
"epoch": 0.84,
"grad_norm": 0.3197622001171112,
"learning_rate": 3.13259214987912e-05,
"loss": 1.2727,
"step": 17460
},
{
"epoch": 0.84,
"grad_norm": 0.3862283229827881,
"learning_rate": 3.130768415805207e-05,
"loss": 1.3147,
"step": 17470
},
{
"epoch": 0.84,
"grad_norm": 0.32670193910598755,
"learning_rate": 3.1289443231692185e-05,
"loss": 1.2828,
"step": 17480
},
{
"epoch": 0.84,
"grad_norm": 0.5309838652610779,
"learning_rate": 3.1271198730080656e-05,
"loss": 1.3715,
"step": 17490
},
{
"epoch": 0.84,
"grad_norm": 0.44820737838745117,
"learning_rate": 3.125295066358861e-05,
"loss": 1.368,
"step": 17500
},
{
"epoch": 0.84,
"grad_norm": 0.3359789550304413,
"learning_rate": 3.1234699042589235e-05,
"loss": 1.3904,
"step": 17510
},
{
"epoch": 0.84,
"grad_norm": 0.5568551421165466,
"learning_rate": 3.121644387745769e-05,
"loss": 1.3514,
"step": 17520
},
{
"epoch": 0.84,
"grad_norm": 0.38214004039764404,
"learning_rate": 3.1198185178571196e-05,
"loss": 1.2724,
"step": 17530
},
{
"epoch": 0.84,
"grad_norm": 0.3803871273994446,
"learning_rate": 3.1179922956308935e-05,
"loss": 1.3241,
"step": 17540
},
{
"epoch": 0.84,
"grad_norm": 0.4368240237236023,
"learning_rate": 3.1161657221052146e-05,
"loss": 1.3301,
"step": 17550
},
{
"epoch": 0.84,
"grad_norm": 0.40888023376464844,
"learning_rate": 3.114338798318402e-05,
"loss": 1.3028,
"step": 17560
},
{
"epoch": 0.84,
"grad_norm": 0.3746178150177002,
"learning_rate": 3.1125115253089764e-05,
"loss": 1.314,
"step": 17570
},
{
"epoch": 0.84,
"grad_norm": 0.4408790171146393,
"learning_rate": 3.110683904115657e-05,
"loss": 1.3684,
"step": 17580
},
{
"epoch": 0.84,
"grad_norm": 0.32696792483329773,
"learning_rate": 3.108855935777359e-05,
"loss": 1.3744,
"step": 17590
},
{
"epoch": 0.84,
"grad_norm": 0.37197616696357727,
"learning_rate": 3.107027621333196e-05,
"loss": 1.3104,
"step": 17600
},
{
"epoch": 0.85,
"grad_norm": 0.34846165776252747,
"learning_rate": 3.1051989618224794e-05,
"loss": 1.3187,
"step": 17610
},
{
"epoch": 0.85,
"grad_norm": 0.5278374552726746,
"learning_rate": 3.103369958284714e-05,
"loss": 1.3431,
"step": 17620
},
{
"epoch": 0.85,
"grad_norm": 0.3801672160625458,
"learning_rate": 3.1015406117596055e-05,
"loss": 1.3385,
"step": 17630
},
{
"epoch": 0.85,
"grad_norm": 0.4286508858203888,
"learning_rate": 3.099710923287047e-05,
"loss": 1.3018,
"step": 17640
},
{
"epoch": 0.85,
"grad_norm": 0.48649460077285767,
"learning_rate": 3.097880893907133e-05,
"loss": 1.4465,
"step": 17650
},
{
"epoch": 0.85,
"grad_norm": 0.4192192852497101,
"learning_rate": 3.0960505246601476e-05,
"loss": 1.3427,
"step": 17660
},
{
"epoch": 0.85,
"grad_norm": 0.36535248160362244,
"learning_rate": 3.094219816586569e-05,
"loss": 1.2787,
"step": 17670
},
{
"epoch": 0.85,
"grad_norm": 0.4469853937625885,
"learning_rate": 3.0923887707270697e-05,
"loss": 1.2635,
"step": 17680
},
{
"epoch": 0.85,
"grad_norm": 0.372471421957016,
"learning_rate": 3.0905573881225117e-05,
"loss": 1.2972,
"step": 17690
},
{
"epoch": 0.85,
"grad_norm": 0.39072540402412415,
"learning_rate": 3.08872566981395e-05,
"loss": 1.2837,
"step": 17700
},
{
"epoch": 0.85,
"grad_norm": 0.38056156039237976,
"learning_rate": 3.0868936168426305e-05,
"loss": 1.439,
"step": 17710
},
{
"epoch": 0.85,
"grad_norm": 0.3827475309371948,
"learning_rate": 3.085061230249987e-05,
"loss": 1.4416,
"step": 17720
},
{
"epoch": 0.85,
"grad_norm": 0.4099487364292145,
"learning_rate": 3.083228511077647e-05,
"loss": 1.2713,
"step": 17730
},
{
"epoch": 0.85,
"grad_norm": 0.4040149450302124,
"learning_rate": 3.0813954603674236e-05,
"loss": 1.3036,
"step": 17740
},
{
"epoch": 0.85,
"grad_norm": 0.37751883268356323,
"learning_rate": 3.0795620791613196e-05,
"loss": 1.3969,
"step": 17750
},
{
"epoch": 0.85,
"grad_norm": 0.3975282907485962,
"learning_rate": 3.077728368501527e-05,
"loss": 1.2581,
"step": 17760
},
{
"epoch": 0.85,
"grad_norm": 0.29947972297668457,
"learning_rate": 3.075894329430421e-05,
"loss": 1.258,
"step": 17770
},
{
"epoch": 0.85,
"grad_norm": 0.3819504678249359,
"learning_rate": 3.07405996299057e-05,
"loss": 1.2929,
"step": 17780
},
{
"epoch": 0.85,
"grad_norm": 0.33974966406822205,
"learning_rate": 3.072408754156254e-05,
"loss": 1.2858,
"step": 17790
},
{
"epoch": 0.85,
"grad_norm": 0.465350478887558,
"learning_rate": 3.0705737685887146e-05,
"loss": 1.3202,
"step": 17800
},
{
"epoch": 0.85,
"grad_norm": 0.43890687823295593,
"learning_rate": 3.0687384586769155e-05,
"loss": 1.3268,
"step": 17810
},
{
"epoch": 0.86,
"grad_norm": 0.4028676748275757,
"learning_rate": 3.066902825464145e-05,
"loss": 1.3434,
"step": 17820
},
{
"epoch": 0.86,
"grad_norm": 0.36824607849121094,
"learning_rate": 3.0650668699938724e-05,
"loss": 1.3399,
"step": 17830
},
{
"epoch": 0.86,
"grad_norm": 0.404031902551651,
"learning_rate": 3.063230593309754e-05,
"loss": 1.2937,
"step": 17840
},
{
"epoch": 0.86,
"grad_norm": 0.38140177726745605,
"learning_rate": 3.0613939964556236e-05,
"loss": 1.3424,
"step": 17850
},
{
"epoch": 0.86,
"grad_norm": 0.3252469003200531,
"learning_rate": 3.059557080475501e-05,
"loss": 1.337,
"step": 17860
},
{
"epoch": 0.86,
"grad_norm": 0.3256832957267761,
"learning_rate": 3.0577198464135875e-05,
"loss": 1.3061,
"step": 17870
},
{
"epoch": 0.86,
"grad_norm": 0.310890257358551,
"learning_rate": 3.055882295314262e-05,
"loss": 1.3024,
"step": 17880
},
{
"epoch": 0.86,
"grad_norm": 0.6329560279846191,
"learning_rate": 3.054044428222088e-05,
"loss": 1.3477,
"step": 17890
},
{
"epoch": 0.86,
"grad_norm": 0.41712427139282227,
"learning_rate": 3.052206246181803e-05,
"loss": 1.339,
"step": 17900
},
{
"epoch": 0.86,
"grad_norm": 0.614782989025116,
"learning_rate": 3.050367750238329e-05,
"loss": 1.295,
"step": 17910
},
{
"epoch": 0.86,
"grad_norm": 0.4023633301258087,
"learning_rate": 3.0485289414367634e-05,
"loss": 1.2927,
"step": 17920
},
{
"epoch": 0.86,
"grad_norm": 0.32646626234054565,
"learning_rate": 3.046689820822382e-05,
"loss": 1.2976,
"step": 17930
},
{
"epoch": 0.86,
"grad_norm": 0.4776906371116638,
"learning_rate": 3.04485038944064e-05,
"loss": 1.2978,
"step": 17940
},
{
"epoch": 0.86,
"grad_norm": 0.4123288094997406,
"learning_rate": 3.0430106483371645e-05,
"loss": 1.3075,
"step": 17950
},
{
"epoch": 0.86,
"grad_norm": 0.38605058193206787,
"learning_rate": 3.0411705985577633e-05,
"loss": 1.3494,
"step": 17960
},
{
"epoch": 0.86,
"grad_norm": 0.40366214513778687,
"learning_rate": 3.0393302411484175e-05,
"loss": 1.4061,
"step": 17970
},
{
"epoch": 0.86,
"grad_norm": 0.3741171956062317,
"learning_rate": 3.0374895771552824e-05,
"loss": 1.3206,
"step": 17980
},
{
"epoch": 0.86,
"grad_norm": 0.33325356245040894,
"learning_rate": 3.0356486076246906e-05,
"loss": 1.329,
"step": 17990
},
{
"epoch": 0.86,
"grad_norm": 0.42309334874153137,
"learning_rate": 3.0338073336031448e-05,
"loss": 1.3965,
"step": 18000
},
{
"epoch": 0.86,
"grad_norm": 0.35031577944755554,
"learning_rate": 3.0319657561373227e-05,
"loss": 1.3195,
"step": 18010
},
{
"epoch": 0.86,
"grad_norm": 0.43163833022117615,
"learning_rate": 3.0301238762740753e-05,
"loss": 1.243,
"step": 18020
},
{
"epoch": 0.87,
"grad_norm": 0.3277037441730499,
"learning_rate": 3.028281695060422e-05,
"loss": 1.3218,
"step": 18030
},
{
"epoch": 0.87,
"grad_norm": 0.39240947365760803,
"learning_rate": 3.0264392135435584e-05,
"loss": 1.347,
"step": 18040
},
{
"epoch": 0.87,
"grad_norm": 0.3747977316379547,
"learning_rate": 3.0245964327708465e-05,
"loss": 1.3545,
"step": 18050
},
{
"epoch": 0.87,
"grad_norm": 0.3396255075931549,
"learning_rate": 3.0227533537898212e-05,
"loss": 1.2492,
"step": 18060
},
{
"epoch": 0.87,
"grad_norm": 0.4028255045413971,
"learning_rate": 3.0209099776481858e-05,
"loss": 1.2501,
"step": 18070
},
{
"epoch": 0.87,
"grad_norm": 0.47649094462394714,
"learning_rate": 3.0190663053938117e-05,
"loss": 1.3222,
"step": 18080
},
{
"epoch": 0.87,
"grad_norm": 0.4251900613307953,
"learning_rate": 3.017222338074741e-05,
"loss": 1.3374,
"step": 18090
},
{
"epoch": 0.87,
"grad_norm": 0.40560251474380493,
"learning_rate": 3.0153780767391814e-05,
"loss": 1.4053,
"step": 18100
},
{
"epoch": 0.87,
"grad_norm": 0.3165496289730072,
"learning_rate": 3.0135335224355082e-05,
"loss": 1.4104,
"step": 18110
},
{
"epoch": 0.87,
"grad_norm": 0.3841902017593384,
"learning_rate": 3.0116886762122642e-05,
"loss": 1.3298,
"step": 18120
},
{
"epoch": 0.87,
"grad_norm": 0.5057214498519897,
"learning_rate": 3.0098435391181555e-05,
"loss": 1.4277,
"step": 18130
},
{
"epoch": 0.87,
"grad_norm": 0.4370133578777313,
"learning_rate": 3.007998112202059e-05,
"loss": 1.3288,
"step": 18140
},
{
"epoch": 0.87,
"grad_norm": 0.34874579310417175,
"learning_rate": 3.006152396513009e-05,
"loss": 1.2571,
"step": 18150
},
{
"epoch": 0.87,
"grad_norm": 0.42282629013061523,
"learning_rate": 3.00430639310021e-05,
"loss": 1.3445,
"step": 18160
},
{
"epoch": 0.87,
"grad_norm": 0.4398539364337921,
"learning_rate": 3.002460103013028e-05,
"loss": 1.3409,
"step": 18170
},
{
"epoch": 0.87,
"grad_norm": 0.32258713245391846,
"learning_rate": 3.0006135273009896e-05,
"loss": 1.3378,
"step": 18180
},
{
"epoch": 0.87,
"grad_norm": 0.36342617869377136,
"learning_rate": 2.9987666670137883e-05,
"loss": 1.2794,
"step": 18190
},
{
"epoch": 0.87,
"grad_norm": 0.4183402359485626,
"learning_rate": 2.996919523201276e-05,
"loss": 1.3385,
"step": 18200
},
{
"epoch": 0.87,
"grad_norm": 0.34227970242500305,
"learning_rate": 2.9950720969134667e-05,
"loss": 1.387,
"step": 18210
},
{
"epoch": 0.87,
"grad_norm": 0.4076346158981323,
"learning_rate": 2.993224389200535e-05,
"loss": 1.3299,
"step": 18220
},
{
"epoch": 0.88,
"grad_norm": 0.511390745639801,
"learning_rate": 2.991376401112816e-05,
"loss": 1.3005,
"step": 18230
},
{
"epoch": 0.88,
"grad_norm": 0.3919171988964081,
"learning_rate": 2.9895281337008036e-05,
"loss": 1.3628,
"step": 18240
},
{
"epoch": 0.88,
"grad_norm": 0.3633108139038086,
"learning_rate": 2.9876795880151503e-05,
"loss": 1.2436,
"step": 18250
},
{
"epoch": 0.88,
"grad_norm": 0.5091758966445923,
"learning_rate": 2.9858307651066675e-05,
"loss": 1.3403,
"step": 18260
},
{
"epoch": 0.88,
"grad_norm": 0.38742539286613464,
"learning_rate": 2.983981666026324e-05,
"loss": 1.2684,
"step": 18270
},
{
"epoch": 0.88,
"grad_norm": 0.3175409138202667,
"learning_rate": 2.982132291825243e-05,
"loss": 1.3369,
"step": 18280
},
{
"epoch": 0.88,
"grad_norm": 0.3371163010597229,
"learning_rate": 2.98028264355471e-05,
"loss": 1.3419,
"step": 18290
},
{
"epoch": 0.88,
"grad_norm": 0.40000560879707336,
"learning_rate": 2.9784327222661607e-05,
"loss": 1.3205,
"step": 18300
},
{
"epoch": 0.88,
"grad_norm": 0.3192661702632904,
"learning_rate": 2.9765825290111882e-05,
"loss": 1.3881,
"step": 18310
},
{
"epoch": 0.88,
"grad_norm": 0.3871364891529083,
"learning_rate": 2.9747320648415407e-05,
"loss": 1.318,
"step": 18320
},
{
"epoch": 0.88,
"grad_norm": 0.30260416865348816,
"learning_rate": 2.9728813308091187e-05,
"loss": 1.3148,
"step": 18330
},
{
"epoch": 0.88,
"grad_norm": 0.3304779827594757,
"learning_rate": 2.971030327965978e-05,
"loss": 1.2858,
"step": 18340
},
{
"epoch": 0.88,
"grad_norm": 0.46220386028289795,
"learning_rate": 2.9691790573643264e-05,
"loss": 1.2579,
"step": 18350
},
{
"epoch": 0.88,
"grad_norm": 0.3271459937095642,
"learning_rate": 2.9673275200565233e-05,
"loss": 1.3049,
"step": 18360
},
{
"epoch": 0.88,
"grad_norm": 0.33543136715888977,
"learning_rate": 2.965475717095081e-05,
"loss": 1.3687,
"step": 18370
},
{
"epoch": 0.88,
"grad_norm": 0.35757988691329956,
"learning_rate": 2.9636236495326603e-05,
"loss": 1.3254,
"step": 18380
},
{
"epoch": 0.88,
"grad_norm": 0.3352295458316803,
"learning_rate": 2.961771318422077e-05,
"loss": 1.3339,
"step": 18390
},
{
"epoch": 0.88,
"grad_norm": 0.37242066860198975,
"learning_rate": 2.9599187248162917e-05,
"loss": 1.205,
"step": 18400
},
{
"epoch": 0.88,
"grad_norm": 0.37466487288475037,
"learning_rate": 2.9580658697684177e-05,
"loss": 1.3813,
"step": 18410
},
{
"epoch": 0.88,
"grad_norm": 0.36901089549064636,
"learning_rate": 2.9562127543317154e-05,
"loss": 1.2585,
"step": 18420
},
{
"epoch": 0.88,
"grad_norm": 0.31505072116851807,
"learning_rate": 2.954359379559592e-05,
"loss": 1.3431,
"step": 18430
},
{
"epoch": 0.89,
"grad_norm": 0.38216879963874817,
"learning_rate": 2.9525057465056062e-05,
"loss": 1.2952,
"step": 18440
},
{
"epoch": 0.89,
"grad_norm": 0.3147180378437042,
"learning_rate": 2.9506518562234587e-05,
"loss": 1.3461,
"step": 18450
},
{
"epoch": 0.89,
"grad_norm": 0.3719359338283539,
"learning_rate": 2.9487977097669994e-05,
"loss": 1.3889,
"step": 18460
},
{
"epoch": 0.89,
"grad_norm": 0.3434627056121826,
"learning_rate": 2.9469433081902238e-05,
"loss": 1.3345,
"step": 18470
},
{
"epoch": 0.89,
"grad_norm": 0.34499964118003845,
"learning_rate": 2.94508865254727e-05,
"loss": 1.3782,
"step": 18480
},
{
"epoch": 0.89,
"grad_norm": 0.3397153615951538,
"learning_rate": 2.943233743892424e-05,
"loss": 1.3621,
"step": 18490
},
{
"epoch": 0.89,
"grad_norm": 0.4036625921726227,
"learning_rate": 2.941378583280113e-05,
"loss": 1.2894,
"step": 18500
},
{
"epoch": 0.89,
"grad_norm": 0.35769763588905334,
"learning_rate": 2.9395231717649075e-05,
"loss": 1.2453,
"step": 18510
},
{
"epoch": 0.89,
"grad_norm": 0.42501336336135864,
"learning_rate": 2.9376675104015234e-05,
"loss": 1.2811,
"step": 18520
},
{
"epoch": 0.89,
"grad_norm": 0.4936717450618744,
"learning_rate": 2.935811600244814e-05,
"loss": 1.3903,
"step": 18530
},
{
"epoch": 0.89,
"grad_norm": 0.3595026731491089,
"learning_rate": 2.9339554423497785e-05,
"loss": 1.2424,
"step": 18540
},
{
"epoch": 0.89,
"grad_norm": 0.45079517364501953,
"learning_rate": 2.9322846893000462e-05,
"loss": 1.2968,
"step": 18550
},
{
"epoch": 0.89,
"grad_norm": 0.44875046610832214,
"learning_rate": 2.930428063609212e-05,
"loss": 1.3188,
"step": 18560
},
{
"epoch": 0.89,
"grad_norm": 0.429980993270874,
"learning_rate": 2.9285711932403375e-05,
"loss": 1.2732,
"step": 18570
},
{
"epoch": 0.89,
"grad_norm": 0.40889135003089905,
"learning_rate": 2.9267140792489656e-05,
"loss": 1.3725,
"step": 18580
},
{
"epoch": 0.89,
"grad_norm": 0.3717174232006073,
"learning_rate": 2.9248567226907787e-05,
"loss": 1.3897,
"step": 18590
},
{
"epoch": 0.89,
"grad_norm": 0.3706338405609131,
"learning_rate": 2.922999124621596e-05,
"loss": 1.3096,
"step": 18600
},
{
"epoch": 0.89,
"grad_norm": 0.44802606105804443,
"learning_rate": 2.9211412860973747e-05,
"loss": 1.2562,
"step": 18610
},
{
"epoch": 0.89,
"grad_norm": 0.3351486623287201,
"learning_rate": 2.919283208174208e-05,
"loss": 1.2818,
"step": 18620
},
{
"epoch": 0.89,
"grad_norm": 0.33683258295059204,
"learning_rate": 2.9174248919083258e-05,
"loss": 1.3799,
"step": 18630
},
{
"epoch": 0.89,
"grad_norm": 0.3339396119117737,
"learning_rate": 2.915566338356092e-05,
"loss": 1.3001,
"step": 18640
},
{
"epoch": 0.9,
"grad_norm": 0.3870473802089691,
"learning_rate": 2.9137075485740077e-05,
"loss": 1.3005,
"step": 18650
},
{
"epoch": 0.9,
"grad_norm": 0.46539434790611267,
"learning_rate": 2.9118485236187075e-05,
"loss": 1.2861,
"step": 18660
},
{
"epoch": 0.9,
"grad_norm": 0.5078036189079285,
"learning_rate": 2.909989264546959e-05,
"loss": 1.2475,
"step": 18670
},
{
"epoch": 0.9,
"grad_norm": 0.3963960111141205,
"learning_rate": 2.9081297724156626e-05,
"loss": 1.3809,
"step": 18680
},
{
"epoch": 0.9,
"grad_norm": 0.4158085286617279,
"learning_rate": 2.9062700482818516e-05,
"loss": 1.3305,
"step": 18690
},
{
"epoch": 0.9,
"grad_norm": 0.36323755979537964,
"learning_rate": 2.904410093202693e-05,
"loss": 1.3688,
"step": 18700
},
{
"epoch": 0.9,
"grad_norm": 0.46389803290367126,
"learning_rate": 2.9025499082354817e-05,
"loss": 1.3278,
"step": 18710
},
{
"epoch": 0.9,
"grad_norm": 0.3773520588874817,
"learning_rate": 2.9006894944376473e-05,
"loss": 1.2745,
"step": 18720
},
{
"epoch": 0.9,
"grad_norm": 0.33558306097984314,
"learning_rate": 2.8988288528667444e-05,
"loss": 1.3647,
"step": 18730
},
{
"epoch": 0.9,
"grad_norm": 0.4792706072330475,
"learning_rate": 2.896967984580462e-05,
"loss": 1.2932,
"step": 18740
},
{
"epoch": 0.9,
"grad_norm": 0.39759260416030884,
"learning_rate": 2.8951068906366152e-05,
"loss": 1.2908,
"step": 18750
},
{
"epoch": 0.9,
"grad_norm": 0.5172334313392639,
"learning_rate": 2.8932455720931485e-05,
"loss": 1.3407,
"step": 18760
},
{
"epoch": 0.9,
"grad_norm": 0.32739338278770447,
"learning_rate": 2.891384030008133e-05,
"loss": 1.3006,
"step": 18770
},
{
"epoch": 0.9,
"grad_norm": 0.38373759388923645,
"learning_rate": 2.889522265439768e-05,
"loss": 1.3327,
"step": 18780
},
{
"epoch": 0.9,
"grad_norm": 0.3752039074897766,
"learning_rate": 2.887660279446378e-05,
"loss": 1.3484,
"step": 18790
},
{
"epoch": 0.9,
"grad_norm": 0.4795733392238617,
"learning_rate": 2.885798073086415e-05,
"loss": 1.2609,
"step": 18800
},
{
"epoch": 0.9,
"grad_norm": 0.4558461010456085,
"learning_rate": 2.8839356474184555e-05,
"loss": 1.2961,
"step": 18810
},
{
"epoch": 0.9,
"grad_norm": 0.47130680084228516,
"learning_rate": 2.8820730035012e-05,
"loss": 1.3521,
"step": 18820
},
{
"epoch": 0.9,
"grad_norm": 0.35040250420570374,
"learning_rate": 2.880210142393474e-05,
"loss": 1.369,
"step": 18830
},
{
"epoch": 0.9,
"grad_norm": 0.4999600052833557,
"learning_rate": 2.878347065154226e-05,
"loss": 1.3335,
"step": 18840
},
{
"epoch": 0.9,
"grad_norm": 0.4078916013240814,
"learning_rate": 2.8764837728425276e-05,
"loss": 1.3602,
"step": 18850
},
{
"epoch": 0.91,
"grad_norm": 0.32398587465286255,
"learning_rate": 2.8746202665175727e-05,
"loss": 1.3791,
"step": 18860
},
{
"epoch": 0.91,
"grad_norm": 0.469713419675827,
"learning_rate": 2.8727565472386765e-05,
"loss": 1.3235,
"step": 18870
},
{
"epoch": 0.91,
"grad_norm": 0.3608604967594147,
"learning_rate": 2.8708926160652748e-05,
"loss": 1.3217,
"step": 18880
},
{
"epoch": 0.91,
"grad_norm": 0.4104129672050476,
"learning_rate": 2.8690284740569256e-05,
"loss": 1.2486,
"step": 18890
},
{
"epoch": 0.91,
"grad_norm": 0.5751768350601196,
"learning_rate": 2.8671641222733054e-05,
"loss": 1.3331,
"step": 18900
},
{
"epoch": 0.91,
"grad_norm": 0.34021031856536865,
"learning_rate": 2.8652995617742095e-05,
"loss": 1.3511,
"step": 18910
},
{
"epoch": 0.91,
"grad_norm": 0.4680972993373871,
"learning_rate": 2.8634347936195538e-05,
"loss": 1.3366,
"step": 18920
},
{
"epoch": 0.91,
"grad_norm": 0.45248040556907654,
"learning_rate": 2.8615698188693695e-05,
"loss": 1.3042,
"step": 18930
},
{
"epoch": 0.91,
"grad_norm": 0.3334675431251526,
"learning_rate": 2.8597046385838076e-05,
"loss": 1.294,
"step": 18940
},
{
"epoch": 0.91,
"grad_norm": 0.3789302110671997,
"learning_rate": 2.857839253823135e-05,
"loss": 1.2591,
"step": 18950
},
{
"epoch": 0.91,
"grad_norm": 0.51711505651474,
"learning_rate": 2.855973665647735e-05,
"loss": 1.2567,
"step": 18960
},
{
"epoch": 0.91,
"grad_norm": 0.40147870779037476,
"learning_rate": 2.854107875118106e-05,
"loss": 1.3679,
"step": 18970
},
{
"epoch": 0.91,
"grad_norm": 0.3672814965248108,
"learning_rate": 2.8522418832948628e-05,
"loss": 1.2266,
"step": 18980
},
{
"epoch": 0.91,
"grad_norm": 0.4023880660533905,
"learning_rate": 2.850375691238732e-05,
"loss": 1.2107,
"step": 18990
},
{
"epoch": 0.91,
"grad_norm": 0.37205177545547485,
"learning_rate": 2.8485093000105568e-05,
"loss": 1.378,
"step": 19000
},
{
"epoch": 0.91,
"grad_norm": 0.5134497284889221,
"learning_rate": 2.8466427106712927e-05,
"loss": 1.3105,
"step": 19010
},
{
"epoch": 0.91,
"grad_norm": 0.3608156740665436,
"learning_rate": 2.8447759242820084e-05,
"loss": 1.2457,
"step": 19020
},
{
"epoch": 0.91,
"grad_norm": 0.3240393102169037,
"learning_rate": 2.8429089419038812e-05,
"loss": 1.3519,
"step": 19030
},
{
"epoch": 0.91,
"grad_norm": 0.3641701638698578,
"learning_rate": 2.8410417645982045e-05,
"loss": 1.3694,
"step": 19040
},
{
"epoch": 0.91,
"grad_norm": 0.3733835220336914,
"learning_rate": 2.8391743934263804e-05,
"loss": 1.2754,
"step": 19050
},
{
"epoch": 0.91,
"grad_norm": 0.35240718722343445,
"learning_rate": 2.837306829449921e-05,
"loss": 1.2903,
"step": 19060
},
{
"epoch": 0.92,
"grad_norm": 0.4617578387260437,
"learning_rate": 2.8354390737304488e-05,
"loss": 1.2814,
"step": 19070
},
{
"epoch": 0.92,
"grad_norm": 0.327472984790802,
"learning_rate": 2.8335711273296933e-05,
"loss": 1.2883,
"step": 19080
},
{
"epoch": 0.92,
"grad_norm": 0.49894747138023376,
"learning_rate": 2.831702991309495e-05,
"loss": 1.3621,
"step": 19090
},
{
"epoch": 0.92,
"grad_norm": 0.36389392614364624,
"learning_rate": 2.829834666731801e-05,
"loss": 1.2715,
"step": 19100
},
{
"epoch": 0.92,
"grad_norm": 0.4186304807662964,
"learning_rate": 2.8279661546586654e-05,
"loss": 1.3839,
"step": 19110
},
{
"epoch": 0.92,
"grad_norm": 0.455446720123291,
"learning_rate": 2.8260974561522496e-05,
"loss": 1.2958,
"step": 19120
},
{
"epoch": 0.92,
"grad_norm": 0.5760899186134338,
"learning_rate": 2.8242285722748197e-05,
"loss": 1.3592,
"step": 19130
},
{
"epoch": 0.92,
"grad_norm": 0.36079666018486023,
"learning_rate": 2.8223595040887485e-05,
"loss": 1.2998,
"step": 19140
},
{
"epoch": 0.92,
"grad_norm": 0.3635547459125519,
"learning_rate": 2.820490252656513e-05,
"loss": 1.342,
"step": 19150
},
{
"epoch": 0.92,
"grad_norm": 0.5210075378417969,
"learning_rate": 2.818620819040694e-05,
"loss": 1.3221,
"step": 19160
},
{
"epoch": 0.92,
"grad_norm": 0.3787085711956024,
"learning_rate": 2.8167512043039777e-05,
"loss": 1.3702,
"step": 19170
},
{
"epoch": 0.92,
"grad_norm": 0.4056733548641205,
"learning_rate": 2.8148814095091502e-05,
"loss": 1.334,
"step": 19180
},
{
"epoch": 0.92,
"grad_norm": 0.38574525713920593,
"learning_rate": 2.8130114357191024e-05,
"loss": 1.2298,
"step": 19190
},
{
"epoch": 0.92,
"grad_norm": 0.3561164438724518,
"learning_rate": 2.8111412839968255e-05,
"loss": 1.2726,
"step": 19200
},
{
"epoch": 0.92,
"grad_norm": 0.38271018862724304,
"learning_rate": 2.8092709554054135e-05,
"loss": 1.3057,
"step": 19210
},
{
"epoch": 0.92,
"grad_norm": 0.3936762511730194,
"learning_rate": 2.8074004510080598e-05,
"loss": 1.3538,
"step": 19220
},
{
"epoch": 0.92,
"grad_norm": 0.3254060745239258,
"learning_rate": 2.805529771868056e-05,
"loss": 1.3535,
"step": 19230
},
{
"epoch": 0.92,
"grad_norm": 0.40829741954803467,
"learning_rate": 2.8036589190487978e-05,
"loss": 1.3483,
"step": 19240
},
{
"epoch": 0.92,
"grad_norm": 0.2792168855667114,
"learning_rate": 2.801787893613774e-05,
"loss": 1.2478,
"step": 19250
},
{
"epoch": 0.92,
"grad_norm": 0.40997618436813354,
"learning_rate": 2.7999166966265765e-05,
"loss": 1.3509,
"step": 19260
},
{
"epoch": 0.92,
"grad_norm": 0.3296857178211212,
"learning_rate": 2.798045329150892e-05,
"loss": 1.3005,
"step": 19270
},
{
"epoch": 0.93,
"grad_norm": 0.47090160846710205,
"learning_rate": 2.796173792250503e-05,
"loss": 1.3598,
"step": 19280
},
{
"epoch": 0.93,
"grad_norm": 0.3931998014450073,
"learning_rate": 2.7943020869892918e-05,
"loss": 1.3268,
"step": 19290
},
{
"epoch": 0.93,
"grad_norm": 0.34435802698135376,
"learning_rate": 2.7924302144312332e-05,
"loss": 1.334,
"step": 19300
},
{
"epoch": 0.93,
"grad_norm": 0.4257373511791229,
"learning_rate": 2.7905581756403987e-05,
"loss": 1.3029,
"step": 19310
},
{
"epoch": 0.93,
"grad_norm": 0.4704016149044037,
"learning_rate": 2.7886859716809553e-05,
"loss": 1.3613,
"step": 19320
},
{
"epoch": 0.93,
"grad_norm": 0.326539546251297,
"learning_rate": 2.786813603617161e-05,
"loss": 1.2939,
"step": 19330
},
{
"epoch": 0.93,
"grad_norm": 0.3895120918750763,
"learning_rate": 2.784941072513369e-05,
"loss": 1.3006,
"step": 19340
},
{
"epoch": 0.93,
"grad_norm": 0.44412752985954285,
"learning_rate": 2.7830683794340252e-05,
"loss": 1.3039,
"step": 19350
},
{
"epoch": 0.93,
"grad_norm": 0.34828612208366394,
"learning_rate": 2.7811955254436667e-05,
"loss": 1.3653,
"step": 19360
},
{
"epoch": 0.93,
"grad_norm": 0.4024435877799988,
"learning_rate": 2.779322511606924e-05,
"loss": 1.3929,
"step": 19370
},
{
"epoch": 0.93,
"grad_norm": 0.3219032287597656,
"learning_rate": 2.7774493389885146e-05,
"loss": 1.3607,
"step": 19380
},
{
"epoch": 0.93,
"grad_norm": 0.3581404387950897,
"learning_rate": 2.7755760086532513e-05,
"loss": 1.295,
"step": 19390
},
{
"epoch": 0.93,
"grad_norm": 0.38259002566337585,
"learning_rate": 2.773702521666033e-05,
"loss": 1.339,
"step": 19400
},
{
"epoch": 0.93,
"grad_norm": 0.3654376268386841,
"learning_rate": 2.7718288790918477e-05,
"loss": 1.3302,
"step": 19410
},
{
"epoch": 0.93,
"grad_norm": 0.3577955663204193,
"learning_rate": 2.7699550819957744e-05,
"loss": 1.2844,
"step": 19420
},
{
"epoch": 0.93,
"grad_norm": 0.3431689739227295,
"learning_rate": 2.768081131442977e-05,
"loss": 1.3271,
"step": 19430
},
{
"epoch": 0.93,
"grad_norm": 0.47660794854164124,
"learning_rate": 2.7662070284987086e-05,
"loss": 1.3906,
"step": 19440
},
{
"epoch": 0.93,
"grad_norm": 0.3448266088962555,
"learning_rate": 2.764332774228308e-05,
"loss": 1.2866,
"step": 19450
},
{
"epoch": 0.93,
"grad_norm": 0.39191773533821106,
"learning_rate": 2.7624583696972007e-05,
"loss": 1.369,
"step": 19460
},
{
"epoch": 0.93,
"grad_norm": 0.3026762902736664,
"learning_rate": 2.7605838159708973e-05,
"loss": 1.3097,
"step": 19470
},
{
"epoch": 0.94,
"grad_norm": 0.3884013891220093,
"learning_rate": 2.7587091141149924e-05,
"loss": 1.2629,
"step": 19480
},
{
"epoch": 0.94,
"grad_norm": 0.560519814491272,
"learning_rate": 2.7568342651951657e-05,
"loss": 1.4048,
"step": 19490
},
{
"epoch": 0.94,
"grad_norm": 0.46949267387390137,
"learning_rate": 2.7549592702771804e-05,
"loss": 1.3319,
"step": 19500
},
{
"epoch": 0.94,
"grad_norm": 0.4962778687477112,
"learning_rate": 2.7530841304268827e-05,
"loss": 1.3255,
"step": 19510
},
{
"epoch": 0.94,
"grad_norm": 0.3905600905418396,
"learning_rate": 2.7512088467102022e-05,
"loss": 1.3372,
"step": 19520
},
{
"epoch": 0.94,
"grad_norm": 0.3942488133907318,
"learning_rate": 2.7493334201931468e-05,
"loss": 1.329,
"step": 19530
},
{
"epoch": 0.94,
"grad_norm": 0.36593902111053467,
"learning_rate": 2.7474578519418102e-05,
"loss": 1.382,
"step": 19540
},
{
"epoch": 0.94,
"grad_norm": 0.43178728222846985,
"learning_rate": 2.7455821430223633e-05,
"loss": 1.3647,
"step": 19550
},
{
"epoch": 0.94,
"grad_norm": 0.4393153786659241,
"learning_rate": 2.7437062945010578e-05,
"loss": 1.3728,
"step": 19560
},
{
"epoch": 0.94,
"grad_norm": 0.39975598454475403,
"learning_rate": 2.7418303074442263e-05,
"loss": 1.3056,
"step": 19570
},
{
"epoch": 0.94,
"grad_norm": 0.34918275475502014,
"learning_rate": 2.7399541829182785e-05,
"loss": 1.3507,
"step": 19580
},
{
"epoch": 0.94,
"grad_norm": 0.5777868032455444,
"learning_rate": 2.7380779219897014e-05,
"loss": 1.3154,
"step": 19590
},
{
"epoch": 0.94,
"grad_norm": 0.3584270477294922,
"learning_rate": 2.7362015257250622e-05,
"loss": 1.3072,
"step": 19600
},
{
"epoch": 0.94,
"grad_norm": 0.367072194814682,
"learning_rate": 2.7343249951910026e-05,
"loss": 1.2678,
"step": 19610
},
{
"epoch": 0.94,
"grad_norm": 0.3917929232120514,
"learning_rate": 2.732448331454243e-05,
"loss": 1.2956,
"step": 19620
},
{
"epoch": 0.94,
"grad_norm": 0.3910132348537445,
"learning_rate": 2.7305715355815776e-05,
"loss": 1.2667,
"step": 19630
},
{
"epoch": 0.94,
"grad_norm": 0.39063411951065063,
"learning_rate": 2.728694608639875e-05,
"loss": 1.3413,
"step": 19640
},
{
"epoch": 0.94,
"grad_norm": 0.4007227420806885,
"learning_rate": 2.7268175516960816e-05,
"loss": 1.2869,
"step": 19650
},
{
"epoch": 0.94,
"grad_norm": 0.37359800934791565,
"learning_rate": 2.724940365817214e-05,
"loss": 1.2417,
"step": 19660
},
{
"epoch": 0.94,
"grad_norm": 0.44577348232269287,
"learning_rate": 2.7230630520703647e-05,
"loss": 1.3954,
"step": 19670
},
{
"epoch": 0.94,
"grad_norm": 0.3215051591396332,
"learning_rate": 2.721185611522698e-05,
"loss": 1.2669,
"step": 19680
},
{
"epoch": 0.95,
"grad_norm": 0.3514328598976135,
"learning_rate": 2.7193080452414498e-05,
"loss": 1.3426,
"step": 19690
},
{
"epoch": 0.95,
"grad_norm": 0.34993141889572144,
"learning_rate": 2.717430354293928e-05,
"loss": 1.3058,
"step": 19700
},
{
"epoch": 0.95,
"grad_norm": 0.38943201303482056,
"learning_rate": 2.7155525397475113e-05,
"loss": 1.3078,
"step": 19710
},
{
"epoch": 0.95,
"grad_norm": 0.39633411169052124,
"learning_rate": 2.7136746026696496e-05,
"loss": 1.2633,
"step": 19720
},
{
"epoch": 0.95,
"grad_norm": 0.37116166949272156,
"learning_rate": 2.7117965441278596e-05,
"loss": 1.414,
"step": 19730
},
{
"epoch": 0.95,
"grad_norm": 0.31941166520118713,
"learning_rate": 2.70991836518973e-05,
"loss": 1.3716,
"step": 19740
},
{
"epoch": 0.95,
"grad_norm": 0.3752063810825348,
"learning_rate": 2.7080400669229177e-05,
"loss": 1.3609,
"step": 19750
},
{
"epoch": 0.95,
"grad_norm": 0.40461257100105286,
"learning_rate": 2.7061616503951453e-05,
"loss": 1.4751,
"step": 19760
},
{
"epoch": 0.95,
"grad_norm": 0.3740611970424652,
"learning_rate": 2.704283116674205e-05,
"loss": 1.3955,
"step": 19770
},
{
"epoch": 0.95,
"grad_norm": 0.3295096755027771,
"learning_rate": 2.7024044668279537e-05,
"loss": 1.2757,
"step": 19780
},
{
"epoch": 0.95,
"grad_norm": 0.4445335865020752,
"learning_rate": 2.700525701924315e-05,
"loss": 1.2646,
"step": 19790
},
{
"epoch": 0.95,
"grad_norm": 0.3357381820678711,
"learning_rate": 2.698646823031279e-05,
"loss": 1.3725,
"step": 19800
},
{
"epoch": 0.95,
"grad_norm": 0.5198288559913635,
"learning_rate": 2.6967678312168988e-05,
"loss": 1.3176,
"step": 19810
},
{
"epoch": 0.95,
"grad_norm": 0.3567744791507721,
"learning_rate": 2.694888727549294e-05,
"loss": 1.281,
"step": 19820
},
{
"epoch": 0.95,
"grad_norm": 0.3883604109287262,
"learning_rate": 2.693009513096645e-05,
"loss": 1.316,
"step": 19830
},
{
"epoch": 0.95,
"grad_norm": 0.4114549458026886,
"learning_rate": 2.691130188927196e-05,
"loss": 1.316,
"step": 19840
},
{
"epoch": 0.95,
"grad_norm": 0.4282682240009308,
"learning_rate": 2.6892507561092557e-05,
"loss": 1.4003,
"step": 19850
},
{
"epoch": 0.95,
"grad_norm": 0.35041624307632446,
"learning_rate": 2.687371215711192e-05,
"loss": 1.2724,
"step": 19860
},
{
"epoch": 0.95,
"grad_norm": 0.34613433480262756,
"learning_rate": 2.685491568801436e-05,
"loss": 1.2403,
"step": 19870
},
{
"epoch": 0.95,
"grad_norm": 0.4286148250102997,
"learning_rate": 2.6836118164484764e-05,
"loss": 1.2417,
"step": 19880
},
{
"epoch": 0.95,
"grad_norm": 0.36300161480903625,
"learning_rate": 2.6819199500600307e-05,
"loss": 1.2583,
"step": 19890
},
{
"epoch": 0.96,
"grad_norm": 0.4007517099380493,
"learning_rate": 2.6800400003088914e-05,
"loss": 1.2994,
"step": 19900
},
{
"epoch": 0.96,
"grad_norm": 0.5609369277954102,
"learning_rate": 2.67815994821351e-05,
"loss": 1.3181,
"step": 19910
},
{
"epoch": 0.96,
"grad_norm": 0.29741737246513367,
"learning_rate": 2.6762797948426066e-05,
"loss": 1.3903,
"step": 19920
},
{
"epoch": 0.96,
"grad_norm": 0.3621099591255188,
"learning_rate": 2.674399541264958e-05,
"loss": 1.3438,
"step": 19930
},
{
"epoch": 0.96,
"grad_norm": 0.5386543869972229,
"learning_rate": 2.6725191885494018e-05,
"loss": 1.2711,
"step": 19940
},
{
"epoch": 0.96,
"grad_norm": 0.36603888869285583,
"learning_rate": 2.670638737764829e-05,
"loss": 1.2758,
"step": 19950
},
{
"epoch": 0.96,
"grad_norm": 0.3907409608364105,
"learning_rate": 2.6687581899801866e-05,
"loss": 1.3185,
"step": 19960
},
{
"epoch": 0.96,
"grad_norm": 0.3582545220851898,
"learning_rate": 2.6668775462644784e-05,
"loss": 1.3659,
"step": 19970
},
{
"epoch": 0.96,
"grad_norm": 0.38071659207344055,
"learning_rate": 2.6649968076867603e-05,
"loss": 1.2488,
"step": 19980
},
{
"epoch": 0.96,
"grad_norm": 0.3519297242164612,
"learning_rate": 2.6631159753161444e-05,
"loss": 1.4007,
"step": 19990
},
{
"epoch": 0.96,
"grad_norm": 0.34724199771881104,
"learning_rate": 2.6612350502217946e-05,
"loss": 1.2946,
"step": 20000
}
],
"logging_steps": 10,
"max_steps": 41668,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 2500,
"total_flos": 2.7454340436831764e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}