Qwen2-VL-2B-Instruct-SFT / trainer_state.json
zzzKAI's picture
Model save
8de82f3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9994440473353983,
"eval_steps": 500,
"global_step": 1573,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0031768723691525695,
"grad_norm": 17.809675216674805,
"learning_rate": 6.329113924050634e-07,
"loss": 1.9316,
"step": 5
},
{
"epoch": 0.006353744738305139,
"grad_norm": 14.445319175720215,
"learning_rate": 1.2658227848101267e-06,
"loss": 1.9416,
"step": 10
},
{
"epoch": 0.009530617107457708,
"grad_norm": 8.333172798156738,
"learning_rate": 1.8987341772151901e-06,
"loss": 1.79,
"step": 15
},
{
"epoch": 0.012707489476610278,
"grad_norm": 5.169636249542236,
"learning_rate": 2.5316455696202535e-06,
"loss": 1.5866,
"step": 20
},
{
"epoch": 0.015884361845762845,
"grad_norm": 4.046501636505127,
"learning_rate": 3.164556962025317e-06,
"loss": 1.3408,
"step": 25
},
{
"epoch": 0.019061234214915415,
"grad_norm": 5.918606758117676,
"learning_rate": 3.7974683544303802e-06,
"loss": 1.0916,
"step": 30
},
{
"epoch": 0.022238106584067985,
"grad_norm": 3.1505024433135986,
"learning_rate": 4.430379746835443e-06,
"loss": 1.0186,
"step": 35
},
{
"epoch": 0.025414978953220556,
"grad_norm": 3.3473265171051025,
"learning_rate": 5.063291139240507e-06,
"loss": 0.9224,
"step": 40
},
{
"epoch": 0.028591851322373123,
"grad_norm": 3.4320054054260254,
"learning_rate": 5.69620253164557e-06,
"loss": 0.8603,
"step": 45
},
{
"epoch": 0.03176872369152569,
"grad_norm": 3.0531117916107178,
"learning_rate": 6.329113924050634e-06,
"loss": 0.8021,
"step": 50
},
{
"epoch": 0.03494559606067826,
"grad_norm": 2.459690570831299,
"learning_rate": 6.962025316455697e-06,
"loss": 0.7707,
"step": 55
},
{
"epoch": 0.03812246842983083,
"grad_norm": 2.4381773471832275,
"learning_rate": 7.5949367088607605e-06,
"loss": 0.7425,
"step": 60
},
{
"epoch": 0.041299340798983404,
"grad_norm": 2.486311435699463,
"learning_rate": 8.227848101265824e-06,
"loss": 0.7261,
"step": 65
},
{
"epoch": 0.04447621316813597,
"grad_norm": 3.9224603176116943,
"learning_rate": 8.860759493670886e-06,
"loss": 0.6954,
"step": 70
},
{
"epoch": 0.04765308553728854,
"grad_norm": 15.483328819274902,
"learning_rate": 9.49367088607595e-06,
"loss": 0.6716,
"step": 75
},
{
"epoch": 0.05082995790644111,
"grad_norm": 3.13858962059021,
"learning_rate": 1.0126582278481014e-05,
"loss": 0.6977,
"step": 80
},
{
"epoch": 0.05400683027559368,
"grad_norm": 2.8523001670837402,
"learning_rate": 1.0759493670886076e-05,
"loss": 0.6585,
"step": 85
},
{
"epoch": 0.057183702644746245,
"grad_norm": 3.508849859237671,
"learning_rate": 1.139240506329114e-05,
"loss": 0.6787,
"step": 90
},
{
"epoch": 0.06036057501389882,
"grad_norm": 2.026244640350342,
"learning_rate": 1.2025316455696203e-05,
"loss": 0.6398,
"step": 95
},
{
"epoch": 0.06353744738305138,
"grad_norm": 3.702071189880371,
"learning_rate": 1.2658227848101268e-05,
"loss": 0.6447,
"step": 100
},
{
"epoch": 0.06671431975220396,
"grad_norm": 2.9863617420196533,
"learning_rate": 1.329113924050633e-05,
"loss": 0.6313,
"step": 105
},
{
"epoch": 0.06989119212135653,
"grad_norm": 2.7040019035339355,
"learning_rate": 1.3924050632911395e-05,
"loss": 0.6274,
"step": 110
},
{
"epoch": 0.0730680644905091,
"grad_norm": 2.6223275661468506,
"learning_rate": 1.4556962025316457e-05,
"loss": 0.6421,
"step": 115
},
{
"epoch": 0.07624493685966166,
"grad_norm": 2.3697361946105957,
"learning_rate": 1.5189873417721521e-05,
"loss": 0.6268,
"step": 120
},
{
"epoch": 0.07942180922881423,
"grad_norm": 2.227189302444458,
"learning_rate": 1.5822784810126583e-05,
"loss": 0.6415,
"step": 125
},
{
"epoch": 0.08259868159796681,
"grad_norm": 2.316190242767334,
"learning_rate": 1.6455696202531647e-05,
"loss": 0.6076,
"step": 130
},
{
"epoch": 0.08577555396711938,
"grad_norm": 2.2983508110046387,
"learning_rate": 1.708860759493671e-05,
"loss": 0.5924,
"step": 135
},
{
"epoch": 0.08895242633627194,
"grad_norm": 1.7339043617248535,
"learning_rate": 1.7721518987341772e-05,
"loss": 0.6322,
"step": 140
},
{
"epoch": 0.09212929870542451,
"grad_norm": 1.9135723114013672,
"learning_rate": 1.8354430379746836e-05,
"loss": 0.6169,
"step": 145
},
{
"epoch": 0.09530617107457708,
"grad_norm": 1.901855230331421,
"learning_rate": 1.89873417721519e-05,
"loss": 0.6093,
"step": 150
},
{
"epoch": 0.09848304344372964,
"grad_norm": 1.660704493522644,
"learning_rate": 1.9620253164556964e-05,
"loss": 0.6128,
"step": 155
},
{
"epoch": 0.10165991581288222,
"grad_norm": 1.8829810619354248,
"learning_rate": 1.9999901413795314e-05,
"loss": 0.6008,
"step": 160
},
{
"epoch": 0.10483678818203479,
"grad_norm": 2.084669351577759,
"learning_rate": 1.9998792341316304e-05,
"loss": 0.618,
"step": 165
},
{
"epoch": 0.10801366055118736,
"grad_norm": 1.7610174417495728,
"learning_rate": 1.9996451100730896e-05,
"loss": 0.6038,
"step": 170
},
{
"epoch": 0.11119053292033992,
"grad_norm": 2.0930941104888916,
"learning_rate": 1.999287798055413e-05,
"loss": 0.5973,
"step": 175
},
{
"epoch": 0.11436740528949249,
"grad_norm": 2.1560468673706055,
"learning_rate": 1.9988073421107646e-05,
"loss": 0.5826,
"step": 180
},
{
"epoch": 0.11754427765864506,
"grad_norm": 1.6942908763885498,
"learning_rate": 1.998203801446545e-05,
"loss": 0.5689,
"step": 185
},
{
"epoch": 0.12072115002779764,
"grad_norm": 1.6405870914459229,
"learning_rate": 1.9974772504380907e-05,
"loss": 0.5516,
"step": 190
},
{
"epoch": 0.1238980223969502,
"grad_norm": 1.5230438709259033,
"learning_rate": 1.9966277786195137e-05,
"loss": 0.5703,
"step": 195
},
{
"epoch": 0.12707489476610276,
"grad_norm": 1.5323349237442017,
"learning_rate": 1.9956554906726627e-05,
"loss": 0.5821,
"step": 200
},
{
"epoch": 0.13025176713525535,
"grad_norm": 1.4441514015197754,
"learning_rate": 1.994560506414229e-05,
"loss": 0.6008,
"step": 205
},
{
"epoch": 0.13342863950440792,
"grad_norm": 1.5659058094024658,
"learning_rate": 1.9933429607809746e-05,
"loss": 0.586,
"step": 210
},
{
"epoch": 0.1366055118735605,
"grad_norm": 1.8428056240081787,
"learning_rate": 1.9920030038131104e-05,
"loss": 0.5608,
"step": 215
},
{
"epoch": 0.13978238424271305,
"grad_norm": 1.4605518579483032,
"learning_rate": 1.990540800635801e-05,
"loss": 0.5217,
"step": 220
},
{
"epoch": 0.14295925661186562,
"grad_norm": 1.4116017818450928,
"learning_rate": 1.98895653143882e-05,
"loss": 0.5979,
"step": 225
},
{
"epoch": 0.1461361289810182,
"grad_norm": 1.426926612854004,
"learning_rate": 1.9872503914543416e-05,
"loss": 0.5739,
"step": 230
},
{
"epoch": 0.14931300135017075,
"grad_norm": 1.7036380767822266,
"learning_rate": 1.9854225909328845e-05,
"loss": 0.56,
"step": 235
},
{
"epoch": 0.15248987371932332,
"grad_norm": 1.400126576423645,
"learning_rate": 1.9834733551174014e-05,
"loss": 0.5589,
"step": 240
},
{
"epoch": 0.1556667460884759,
"grad_norm": 1.3639277219772339,
"learning_rate": 1.9814029242155217e-05,
"loss": 0.5827,
"step": 245
},
{
"epoch": 0.15884361845762845,
"grad_norm": 1.3997248411178589,
"learning_rate": 1.9792115533699493e-05,
"loss": 0.5768,
"step": 250
},
{
"epoch": 0.16202049082678102,
"grad_norm": 1.3513110876083374,
"learning_rate": 1.976899512627024e-05,
"loss": 0.56,
"step": 255
},
{
"epoch": 0.16519736319593362,
"grad_norm": 1.338847279548645,
"learning_rate": 1.9744670869034407e-05,
"loss": 0.5212,
"step": 260
},
{
"epoch": 0.16837423556508618,
"grad_norm": 1.4440042972564697,
"learning_rate": 1.971914575951138e-05,
"loss": 0.5708,
"step": 265
},
{
"epoch": 0.17155110793423875,
"grad_norm": 1.3869779109954834,
"learning_rate": 1.969242294320362e-05,
"loss": 0.5483,
"step": 270
},
{
"epoch": 0.17472798030339132,
"grad_norm": 1.5152400732040405,
"learning_rate": 1.9664505713209017e-05,
"loss": 0.554,
"step": 275
},
{
"epoch": 0.17790485267254388,
"grad_norm": 1.3730214834213257,
"learning_rate": 1.9635397509815087e-05,
"loss": 0.5478,
"step": 280
},
{
"epoch": 0.18108172504169645,
"grad_norm": 1.4039232730865479,
"learning_rate": 1.9605101920075003e-05,
"loss": 0.5253,
"step": 285
},
{
"epoch": 0.18425859741084902,
"grad_norm": 1.5447697639465332,
"learning_rate": 1.9573622677365572e-05,
"loss": 0.5343,
"step": 290
},
{
"epoch": 0.18743546978000158,
"grad_norm": 1.3592406511306763,
"learning_rate": 1.954096366092717e-05,
"loss": 0.572,
"step": 295
},
{
"epoch": 0.19061234214915415,
"grad_norm": 1.4715423583984375,
"learning_rate": 1.9507128895385676e-05,
"loss": 0.5579,
"step": 300
},
{
"epoch": 0.19378921451830672,
"grad_norm": 1.3932384252548218,
"learning_rate": 1.9472122550256516e-05,
"loss": 0.5669,
"step": 305
},
{
"epoch": 0.19696608688745929,
"grad_norm": 1.216783046722412,
"learning_rate": 1.9435948939430868e-05,
"loss": 0.5204,
"step": 310
},
{
"epoch": 0.20014295925661185,
"grad_norm": 1.2733324766159058,
"learning_rate": 1.9398612520644028e-05,
"loss": 0.5181,
"step": 315
},
{
"epoch": 0.20331983162576445,
"grad_norm": 1.4451696872711182,
"learning_rate": 1.936011789492609e-05,
"loss": 0.5507,
"step": 320
},
{
"epoch": 0.206496703994917,
"grad_norm": 1.3773140907287598,
"learning_rate": 1.9320469806034937e-05,
"loss": 0.5342,
"step": 325
},
{
"epoch": 0.20967357636406958,
"grad_norm": 1.330608606338501,
"learning_rate": 1.9279673139871686e-05,
"loss": 0.5355,
"step": 330
},
{
"epoch": 0.21285044873322215,
"grad_norm": 1.1820790767669678,
"learning_rate": 1.923773292387857e-05,
"loss": 0.5492,
"step": 335
},
{
"epoch": 0.21602732110237471,
"grad_norm": 1.4723114967346191,
"learning_rate": 1.9194654326419405e-05,
"loss": 0.5303,
"step": 340
},
{
"epoch": 0.21920419347152728,
"grad_norm": 1.3733022212982178,
"learning_rate": 1.9150442656142675e-05,
"loss": 0.5085,
"step": 345
},
{
"epoch": 0.22238106584067985,
"grad_norm": 1.2373732328414917,
"learning_rate": 1.910510336132735e-05,
"loss": 0.5333,
"step": 350
},
{
"epoch": 0.22555793820983241,
"grad_norm": 1.3679918050765991,
"learning_rate": 1.9058642029211492e-05,
"loss": 0.5518,
"step": 355
},
{
"epoch": 0.22873481057898498,
"grad_norm": 1.4029483795166016,
"learning_rate": 1.9011064385303697e-05,
"loss": 0.5217,
"step": 360
},
{
"epoch": 0.23191168294813755,
"grad_norm": 1.2620315551757812,
"learning_rate": 1.896237629267757e-05,
"loss": 0.5386,
"step": 365
},
{
"epoch": 0.23508855531729012,
"grad_norm": 1.207796335220337,
"learning_rate": 1.8912583751249184e-05,
"loss": 0.5053,
"step": 370
},
{
"epoch": 0.2382654276864427,
"grad_norm": 1.3968982696533203,
"learning_rate": 1.886169289703771e-05,
"loss": 0.5457,
"step": 375
},
{
"epoch": 0.24144230005559528,
"grad_norm": 1.2062251567840576,
"learning_rate": 1.880971000140926e-05,
"loss": 0.5248,
"step": 380
},
{
"epoch": 0.24461917242474784,
"grad_norm": 1.5815231800079346,
"learning_rate": 1.875664147030406e-05,
"loss": 0.534,
"step": 385
},
{
"epoch": 0.2477960447939004,
"grad_norm": 1.2773513793945312,
"learning_rate": 1.870249384344704e-05,
"loss": 0.5188,
"step": 390
},
{
"epoch": 0.250972917163053,
"grad_norm": 1.4699748754501343,
"learning_rate": 1.864727379354191e-05,
"loss": 0.5271,
"step": 395
},
{
"epoch": 0.2541497895322055,
"grad_norm": 1.5354286432266235,
"learning_rate": 1.8590988125448906e-05,
"loss": 0.5087,
"step": 400
},
{
"epoch": 0.2573266619013581,
"grad_norm": 1.1777474880218506,
"learning_rate": 1.853364377534618e-05,
"loss": 0.4959,
"step": 405
},
{
"epoch": 0.2605035342705107,
"grad_norm": 1.197090744972229,
"learning_rate": 1.8475247809875076e-05,
"loss": 0.5336,
"step": 410
},
{
"epoch": 0.26368040663966325,
"grad_norm": 1.1266974210739136,
"learning_rate": 1.8415807425269273e-05,
"loss": 0.5502,
"step": 415
},
{
"epoch": 0.26685727900881584,
"grad_norm": 1.2330613136291504,
"learning_rate": 1.8355329946467984e-05,
"loss": 0.5332,
"step": 420
},
{
"epoch": 0.2700341513779684,
"grad_norm": 1.227715015411377,
"learning_rate": 1.8293822826213302e-05,
"loss": 0.51,
"step": 425
},
{
"epoch": 0.273211023747121,
"grad_norm": 1.3449698686599731,
"learning_rate": 1.823129364413178e-05,
"loss": 0.5039,
"step": 430
},
{
"epoch": 0.2763878961162735,
"grad_norm": 1.2230104207992554,
"learning_rate": 1.8167750105800367e-05,
"loss": 0.5356,
"step": 435
},
{
"epoch": 0.2795647684854261,
"grad_norm": 1.2294344902038574,
"learning_rate": 1.8103200041796864e-05,
"loss": 0.5379,
"step": 440
},
{
"epoch": 0.28274164085457865,
"grad_norm": 1.120505690574646,
"learning_rate": 1.8037651406734936e-05,
"loss": 0.5017,
"step": 445
},
{
"epoch": 0.28591851322373124,
"grad_norm": 1.1958080530166626,
"learning_rate": 1.7971112278283836e-05,
"loss": 0.5307,
"step": 450
},
{
"epoch": 0.2890953855928838,
"grad_norm": 1.2974480390548706,
"learning_rate": 1.7903590856173004e-05,
"loss": 0.5287,
"step": 455
},
{
"epoch": 0.2922722579620364,
"grad_norm": 1.2087041139602661,
"learning_rate": 1.7835095461181594e-05,
"loss": 0.4981,
"step": 460
},
{
"epoch": 0.29544913033118897,
"grad_norm": 1.2422888278961182,
"learning_rate": 1.7765634534113075e-05,
"loss": 0.5292,
"step": 465
},
{
"epoch": 0.2986260027003415,
"grad_norm": 1.2754653692245483,
"learning_rate": 1.7695216634755076e-05,
"loss": 0.5334,
"step": 470
},
{
"epoch": 0.3018028750694941,
"grad_norm": 1.4292696714401245,
"learning_rate": 1.762385044082454e-05,
"loss": 0.5223,
"step": 475
},
{
"epoch": 0.30497974743864664,
"grad_norm": 1.1018996238708496,
"learning_rate": 1.7551544746898355e-05,
"loss": 0.5291,
"step": 480
},
{
"epoch": 0.30815661980779924,
"grad_norm": 1.2348871231079102,
"learning_rate": 1.7478308463329584e-05,
"loss": 0.5271,
"step": 485
},
{
"epoch": 0.3113334921769518,
"grad_norm": 1.2747186422348022,
"learning_rate": 1.7404150615149433e-05,
"loss": 0.5171,
"step": 490
},
{
"epoch": 0.31451036454610437,
"grad_norm": 1.1623306274414062,
"learning_rate": 1.7329080340955063e-05,
"loss": 0.512,
"step": 495
},
{
"epoch": 0.3176872369152569,
"grad_norm": 1.3075213432312012,
"learning_rate": 1.725310689178344e-05,
"loss": 0.4984,
"step": 500
},
{
"epoch": 0.3208641092844095,
"grad_norm": 1.14993417263031,
"learning_rate": 1.7176239629971325e-05,
"loss": 0.5118,
"step": 505
},
{
"epoch": 0.32404098165356204,
"grad_norm": 1.3417556285858154,
"learning_rate": 1.709848802800151e-05,
"loss": 0.4961,
"step": 510
},
{
"epoch": 0.32721785402271464,
"grad_norm": 1.2760066986083984,
"learning_rate": 1.7019861667335524e-05,
"loss": 0.4922,
"step": 515
},
{
"epoch": 0.33039472639186723,
"grad_norm": 1.1567506790161133,
"learning_rate": 1.6940370237232887e-05,
"loss": 0.4858,
"step": 520
},
{
"epoch": 0.33357159876101977,
"grad_norm": 1.4701095819473267,
"learning_rate": 1.6860023533557106e-05,
"loss": 0.4964,
"step": 525
},
{
"epoch": 0.33674847113017237,
"grad_norm": 1.2197904586791992,
"learning_rate": 1.677883145756848e-05,
"loss": 0.487,
"step": 530
},
{
"epoch": 0.3399253434993249,
"grad_norm": 1.1546249389648438,
"learning_rate": 1.669680401470398e-05,
"loss": 0.4745,
"step": 535
},
{
"epoch": 0.3431022158684775,
"grad_norm": 1.1630467176437378,
"learning_rate": 1.661395131334425e-05,
"loss": 0.5103,
"step": 540
},
{
"epoch": 0.34627908823763004,
"grad_norm": 1.2125153541564941,
"learning_rate": 1.6530283563567927e-05,
"loss": 0.5193,
"step": 545
},
{
"epoch": 0.34945596060678263,
"grad_norm": 1.4535295963287354,
"learning_rate": 1.644581107589346e-05,
"loss": 0.5122,
"step": 550
},
{
"epoch": 0.3526328329759352,
"grad_norm": 1.1809464693069458,
"learning_rate": 1.636054426000849e-05,
"loss": 0.5062,
"step": 555
},
{
"epoch": 0.35580970534508777,
"grad_norm": 1.2318694591522217,
"learning_rate": 1.6274493623487094e-05,
"loss": 0.4834,
"step": 560
},
{
"epoch": 0.3589865777142403,
"grad_norm": 1.2554454803466797,
"learning_rate": 1.6187669770494867e-05,
"loss": 0.5008,
"step": 565
},
{
"epoch": 0.3621634500833929,
"grad_norm": 1.2687321901321411,
"learning_rate": 1.6100083400482197e-05,
"loss": 0.5019,
"step": 570
},
{
"epoch": 0.3653403224525455,
"grad_norm": 1.148877501487732,
"learning_rate": 1.6011745306865726e-05,
"loss": 0.5094,
"step": 575
},
{
"epoch": 0.36851719482169804,
"grad_norm": 1.1435184478759766,
"learning_rate": 1.5922666375698275e-05,
"loss": 0.513,
"step": 580
},
{
"epoch": 0.37169406719085063,
"grad_norm": 1.151990294456482,
"learning_rate": 1.583285758432732e-05,
"loss": 0.4866,
"step": 585
},
{
"epoch": 0.37487093956000317,
"grad_norm": 1.1544941663742065,
"learning_rate": 1.574233000004226e-05,
"loss": 0.4835,
"step": 590
},
{
"epoch": 0.37804781192915576,
"grad_norm": 1.1831696033477783,
"learning_rate": 1.5651094778710548e-05,
"loss": 0.4941,
"step": 595
},
{
"epoch": 0.3812246842983083,
"grad_norm": 1.1588399410247803,
"learning_rate": 1.5559163163402953e-05,
"loss": 0.4891,
"step": 600
},
{
"epoch": 0.3844015566674609,
"grad_norm": 1.0901782512664795,
"learning_rate": 1.5466546483008057e-05,
"loss": 0.4954,
"step": 605
},
{
"epoch": 0.38757842903661344,
"grad_norm": 1.1416994333267212,
"learning_rate": 1.5373256150836168e-05,
"loss": 0.5006,
"step": 610
},
{
"epoch": 0.39075530140576603,
"grad_norm": 1.2424296140670776,
"learning_rate": 1.527930366321284e-05,
"loss": 0.4867,
"step": 615
},
{
"epoch": 0.39393217377491857,
"grad_norm": 1.2324466705322266,
"learning_rate": 1.5184700598062172e-05,
"loss": 0.5112,
"step": 620
},
{
"epoch": 0.39710904614407116,
"grad_norm": 1.193036675453186,
"learning_rate": 1.508945861348003e-05,
"loss": 0.4907,
"step": 625
},
{
"epoch": 0.4002859185132237,
"grad_norm": 1.25301992893219,
"learning_rate": 1.4993589446297396e-05,
"loss": 0.4811,
"step": 630
},
{
"epoch": 0.4034627908823763,
"grad_norm": 1.3160258531570435,
"learning_rate": 1.4897104910634035e-05,
"loss": 0.4873,
"step": 635
},
{
"epoch": 0.4066396632515289,
"grad_norm": 1.1781983375549316,
"learning_rate": 1.4800016896442606e-05,
"loss": 0.5187,
"step": 640
},
{
"epoch": 0.40981653562068143,
"grad_norm": 1.2555766105651855,
"learning_rate": 1.4702337368043452e-05,
"loss": 0.5007,
"step": 645
},
{
"epoch": 0.412993407989834,
"grad_norm": 1.3068451881408691,
"learning_rate": 1.4604078362650212e-05,
"loss": 0.491,
"step": 650
},
{
"epoch": 0.41617028035898657,
"grad_norm": 1.1431891918182373,
"learning_rate": 1.4505251988886455e-05,
"loss": 0.4887,
"step": 655
},
{
"epoch": 0.41934715272813916,
"grad_norm": 1.1333122253417969,
"learning_rate": 1.4405870425293518e-05,
"loss": 0.489,
"step": 660
},
{
"epoch": 0.4225240250972917,
"grad_norm": 1.1263874769210815,
"learning_rate": 1.430594591882971e-05,
"loss": 0.4832,
"step": 665
},
{
"epoch": 0.4257008974664443,
"grad_norm": 1.0670287609100342,
"learning_rate": 1.420549078336113e-05,
"loss": 0.5105,
"step": 670
},
{
"epoch": 0.42887776983559683,
"grad_norm": 1.1833176612854004,
"learning_rate": 1.4104517398144162e-05,
"loss": 0.5155,
"step": 675
},
{
"epoch": 0.43205464220474943,
"grad_norm": 1.251916766166687,
"learning_rate": 1.40030382063e-05,
"loss": 0.4977,
"step": 680
},
{
"epoch": 0.43523151457390197,
"grad_norm": 1.0563840866088867,
"learning_rate": 1.3901065713281248e-05,
"loss": 0.4783,
"step": 685
},
{
"epoch": 0.43840838694305456,
"grad_norm": 1.0980114936828613,
"learning_rate": 1.3798612485330834e-05,
"loss": 0.4698,
"step": 690
},
{
"epoch": 0.44158525931220716,
"grad_norm": 1.427857518196106,
"learning_rate": 1.3695691147933475e-05,
"loss": 0.5045,
"step": 695
},
{
"epoch": 0.4447621316813597,
"grad_norm": 1.2345526218414307,
"learning_rate": 1.3592314384259809e-05,
"loss": 0.5005,
"step": 700
},
{
"epoch": 0.4479390040505123,
"grad_norm": 1.2202253341674805,
"learning_rate": 1.3488494933603418e-05,
"loss": 0.4712,
"step": 705
},
{
"epoch": 0.45111587641966483,
"grad_norm": 1.1247138977050781,
"learning_rate": 1.3384245589810955e-05,
"loss": 0.4886,
"step": 710
},
{
"epoch": 0.4542927487888174,
"grad_norm": 1.2570569515228271,
"learning_rate": 1.3279579199705537e-05,
"loss": 0.4849,
"step": 715
},
{
"epoch": 0.45746962115796996,
"grad_norm": 1.1821649074554443,
"learning_rate": 1.3174508661503591e-05,
"loss": 0.4773,
"step": 720
},
{
"epoch": 0.46064649352712256,
"grad_norm": 1.215633749961853,
"learning_rate": 1.306904692322541e-05,
"loss": 0.49,
"step": 725
},
{
"epoch": 0.4638233658962751,
"grad_norm": 1.062605857849121,
"learning_rate": 1.2963206981099528e-05,
"loss": 0.4886,
"step": 730
},
{
"epoch": 0.4670002382654277,
"grad_norm": 1.2538633346557617,
"learning_rate": 1.2857001877961181e-05,
"loss": 0.4874,
"step": 735
},
{
"epoch": 0.47017711063458023,
"grad_norm": 1.1429924964904785,
"learning_rate": 1.2750444701645013e-05,
"loss": 0.4954,
"step": 740
},
{
"epoch": 0.4733539830037328,
"grad_norm": 1.1001482009887695,
"learning_rate": 1.264354858337225e-05,
"loss": 0.4913,
"step": 745
},
{
"epoch": 0.4765308553728854,
"grad_norm": 1.1915541887283325,
"learning_rate": 1.25363266961325e-05,
"loss": 0.4925,
"step": 750
},
{
"epoch": 0.47970772774203796,
"grad_norm": 1.0904185771942139,
"learning_rate": 1.242879225306043e-05,
"loss": 0.461,
"step": 755
},
{
"epoch": 0.48288460011119055,
"grad_norm": 1.0332921743392944,
"learning_rate": 1.232095850580751e-05,
"loss": 0.4575,
"step": 760
},
{
"epoch": 0.4860614724803431,
"grad_norm": 1.1236475706100464,
"learning_rate": 1.221283874290894e-05,
"loss": 0.4813,
"step": 765
},
{
"epoch": 0.4892383448494957,
"grad_norm": 1.1378127336502075,
"learning_rate": 1.2104446288146143e-05,
"loss": 0.4574,
"step": 770
},
{
"epoch": 0.4924152172186482,
"grad_norm": 1.0330129861831665,
"learning_rate": 1.1995794498904805e-05,
"loss": 0.4872,
"step": 775
},
{
"epoch": 0.4955920895878008,
"grad_norm": 1.2020204067230225,
"learning_rate": 1.1886896764528837e-05,
"loss": 0.4936,
"step": 780
},
{
"epoch": 0.49876896195695336,
"grad_norm": 1.129347324371338,
"learning_rate": 1.1777766504670397e-05,
"loss": 0.4668,
"step": 785
},
{
"epoch": 0.501945834326106,
"grad_norm": 1.1032615900039673,
"learning_rate": 1.1668417167636143e-05,
"loss": 0.4953,
"step": 790
},
{
"epoch": 0.5051227066952585,
"grad_norm": 1.2174361944198608,
"learning_rate": 1.1558862228729985e-05,
"loss": 0.4856,
"step": 795
},
{
"epoch": 0.508299579064411,
"grad_norm": 1.0304588079452515,
"learning_rate": 1.1449115188592505e-05,
"loss": 0.4603,
"step": 800
},
{
"epoch": 0.5114764514335637,
"grad_norm": 1.1562561988830566,
"learning_rate": 1.1339189571537244e-05,
"loss": 0.4764,
"step": 805
},
{
"epoch": 0.5146533238027162,
"grad_norm": 1.1028969287872314,
"learning_rate": 1.1229098923884065e-05,
"loss": 0.4614,
"step": 810
},
{
"epoch": 0.5178301961718688,
"grad_norm": 1.118483304977417,
"learning_rate": 1.1118856812289856e-05,
"loss": 0.4934,
"step": 815
},
{
"epoch": 0.5210070685410214,
"grad_norm": 1.1935844421386719,
"learning_rate": 1.1008476822076638e-05,
"loss": 0.4737,
"step": 820
},
{
"epoch": 0.524183940910174,
"grad_norm": 1.1401269435882568,
"learning_rate": 1.0897972555557465e-05,
"loss": 0.4918,
"step": 825
},
{
"epoch": 0.5273608132793265,
"grad_norm": 1.0230424404144287,
"learning_rate": 1.0787357630360163e-05,
"loss": 0.462,
"step": 830
},
{
"epoch": 0.530537685648479,
"grad_norm": 1.0743178129196167,
"learning_rate": 1.0676645677749215e-05,
"loss": 0.4995,
"step": 835
},
{
"epoch": 0.5337145580176317,
"grad_norm": 1.170649766921997,
"learning_rate": 1.0565850340945955e-05,
"loss": 0.4951,
"step": 840
},
{
"epoch": 0.5368914303867842,
"grad_norm": 1.068259835243225,
"learning_rate": 1.04549852734473e-05,
"loss": 0.4667,
"step": 845
},
{
"epoch": 0.5400683027559368,
"grad_norm": 1.168420433998108,
"learning_rate": 1.0344064137343187e-05,
"loss": 0.4548,
"step": 850
},
{
"epoch": 0.5432451751250893,
"grad_norm": 1.1319665908813477,
"learning_rate": 1.0233100601632986e-05,
"loss": 0.4744,
"step": 855
},
{
"epoch": 0.546422047494242,
"grad_norm": 1.0778429508209229,
"learning_rate": 1.0122108340541053e-05,
"loss": 0.4674,
"step": 860
},
{
"epoch": 0.5495989198633945,
"grad_norm": 1.0879789590835571,
"learning_rate": 1.0011101031831604e-05,
"loss": 0.4536,
"step": 865
},
{
"epoch": 0.552775792232547,
"grad_norm": 1.0822393894195557,
"learning_rate": 9.90009235512321e-06,
"loss": 0.4662,
"step": 870
},
{
"epoch": 0.5559526646016997,
"grad_norm": 1.0387858152389526,
"learning_rate": 9.789095990203025e-06,
"loss": 0.4438,
"step": 875
},
{
"epoch": 0.5591295369708522,
"grad_norm": 1.1683969497680664,
"learning_rate": 9.678125615340986e-06,
"loss": 0.4834,
"step": 880
},
{
"epoch": 0.5623064093400048,
"grad_norm": 1.0404331684112549,
"learning_rate": 9.567194905604245e-06,
"loss": 0.4651,
"step": 885
},
{
"epoch": 0.5654832817091573,
"grad_norm": 0.9641637802124023,
"learning_rate": 9.456317531171947e-06,
"loss": 0.4808,
"step": 890
},
{
"epoch": 0.5686601540783099,
"grad_norm": 0.9650170207023621,
"learning_rate": 9.345507155650645e-06,
"loss": 0.4903,
"step": 895
},
{
"epoch": 0.5718370264474625,
"grad_norm": 1.0568928718566895,
"learning_rate": 9.234777434390492e-06,
"loss": 0.4553,
"step": 900
},
{
"epoch": 0.575013898816615,
"grad_norm": 1.1297322511672974,
"learning_rate": 9.12414201280248e-06,
"loss": 0.4726,
"step": 905
},
{
"epoch": 0.5781907711857676,
"grad_norm": 1.1371909379959106,
"learning_rate": 9.013614524676907e-06,
"loss": 0.4639,
"step": 910
},
{
"epoch": 0.5813676435549202,
"grad_norm": 1.0769912004470825,
"learning_rate": 8.90320859050323e-06,
"loss": 0.4587,
"step": 915
},
{
"epoch": 0.5845445159240727,
"grad_norm": 1.187941551208496,
"learning_rate": 8.792937815791624e-06,
"loss": 0.4551,
"step": 920
},
{
"epoch": 0.5877213882932253,
"grad_norm": 1.0256364345550537,
"learning_rate": 8.682815789396318e-06,
"loss": 0.4703,
"step": 925
},
{
"epoch": 0.5908982606623779,
"grad_norm": 1.176479458808899,
"learning_rate": 8.57285608184104e-06,
"loss": 0.471,
"step": 930
},
{
"epoch": 0.5940751330315305,
"grad_norm": 0.9811561107635498,
"learning_rate": 8.4630722436467e-06,
"loss": 0.4924,
"step": 935
},
{
"epoch": 0.597252005400683,
"grad_norm": 1.0429517030715942,
"learning_rate": 8.353477803661526e-06,
"loss": 0.4565,
"step": 940
},
{
"epoch": 0.6004288777698356,
"grad_norm": 1.0858372449874878,
"learning_rate": 8.24408626739387e-06,
"loss": 0.4519,
"step": 945
},
{
"epoch": 0.6036057501389882,
"grad_norm": 1.1124157905578613,
"learning_rate": 8.134911115347934e-06,
"loss": 0.4947,
"step": 950
},
{
"epoch": 0.6067826225081407,
"grad_norm": 1.1018826961517334,
"learning_rate": 8.02596580136252e-06,
"loss": 0.4721,
"step": 955
},
{
"epoch": 0.6099594948772933,
"grad_norm": 1.2431491613388062,
"learning_rate": 7.917263750953092e-06,
"loss": 0.4765,
"step": 960
},
{
"epoch": 0.6131363672464458,
"grad_norm": 1.1556727886199951,
"learning_rate": 7.80881835965734e-06,
"loss": 0.4622,
"step": 965
},
{
"epoch": 0.6163132396155985,
"grad_norm": 1.1924182176589966,
"learning_rate": 7.700642991384407e-06,
"loss": 0.484,
"step": 970
},
{
"epoch": 0.619490111984751,
"grad_norm": 1.068928599357605,
"learning_rate": 7.592750976768048e-06,
"loss": 0.4538,
"step": 975
},
{
"epoch": 0.6226669843539036,
"grad_norm": 1.0944935083389282,
"learning_rate": 7.485155611523869e-06,
"loss": 0.4763,
"step": 980
},
{
"epoch": 0.6258438567230562,
"grad_norm": 0.954535961151123,
"learning_rate": 7.377870154810869e-06,
"loss": 0.4719,
"step": 985
},
{
"epoch": 0.6290207290922087,
"grad_norm": 1.132608413696289,
"learning_rate": 7.270907827597487e-06,
"loss": 0.4548,
"step": 990
},
{
"epoch": 0.6321976014613613,
"grad_norm": 1.0536706447601318,
"learning_rate": 7.16428181103238e-06,
"loss": 0.4558,
"step": 995
},
{
"epoch": 0.6353744738305138,
"grad_norm": 1.074761986732483,
"learning_rate": 7.058005244820068e-06,
"loss": 0.4838,
"step": 1000
},
{
"epoch": 0.6385513461996665,
"grad_norm": 1.0573989152908325,
"learning_rate": 6.952091225601713e-06,
"loss": 0.4491,
"step": 1005
},
{
"epoch": 0.641728218568819,
"grad_norm": 1.0783993005752563,
"learning_rate": 6.846552805341194e-06,
"loss": 0.4525,
"step": 1010
},
{
"epoch": 0.6449050909379715,
"grad_norm": 1.0637626647949219,
"learning_rate": 6.7414029897167e-06,
"loss": 0.4771,
"step": 1015
},
{
"epoch": 0.6480819633071241,
"grad_norm": 1.1191637516021729,
"learning_rate": 6.636654736518007e-06,
"loss": 0.4574,
"step": 1020
},
{
"epoch": 0.6512588356762767,
"grad_norm": 1.099088430404663,
"learning_rate": 6.532320954049682e-06,
"loss": 0.4604,
"step": 1025
},
{
"epoch": 0.6544357080454293,
"grad_norm": 1.058421015739441,
"learning_rate": 6.4284144995403565e-06,
"loss": 0.4469,
"step": 1030
},
{
"epoch": 0.6576125804145818,
"grad_norm": 1.0653492212295532,
"learning_rate": 6.324948177558307e-06,
"loss": 0.4554,
"step": 1035
},
{
"epoch": 0.6607894527837345,
"grad_norm": 1.049944519996643,
"learning_rate": 6.2219347384335505e-06,
"loss": 0.4835,
"step": 1040
},
{
"epoch": 0.663966325152887,
"grad_norm": 1.0621883869171143,
"learning_rate": 6.119386876686571e-06,
"loss": 0.463,
"step": 1045
},
{
"epoch": 0.6671431975220395,
"grad_norm": 1.0801547765731812,
"learning_rate": 6.017317229463968e-06,
"loss": 0.4681,
"step": 1050
},
{
"epoch": 0.6703200698911921,
"grad_norm": 1.023689866065979,
"learning_rate": 5.91573837498115e-06,
"loss": 0.4587,
"step": 1055
},
{
"epoch": 0.6734969422603447,
"grad_norm": 1.0890988111495972,
"learning_rate": 5.8146628309723155e-06,
"loss": 0.4363,
"step": 1060
},
{
"epoch": 0.6766738146294973,
"grad_norm": 1.0633463859558105,
"learning_rate": 5.714103053147852e-06,
"loss": 0.4443,
"step": 1065
},
{
"epoch": 0.6798506869986498,
"grad_norm": 1.1121097803115845,
"learning_rate": 5.6140714336594086e-06,
"loss": 0.4563,
"step": 1070
},
{
"epoch": 0.6830275593678024,
"grad_norm": 1.051448106765747,
"learning_rate": 5.514580299572801e-06,
"loss": 0.4478,
"step": 1075
},
{
"epoch": 0.686204431736955,
"grad_norm": 1.077275037765503,
"learning_rate": 5.415641911348893e-06,
"loss": 0.4798,
"step": 1080
},
{
"epoch": 0.6893813041061075,
"grad_norm": 1.1081876754760742,
"learning_rate": 5.31726846133275e-06,
"loss": 0.4511,
"step": 1085
},
{
"epoch": 0.6925581764752601,
"grad_norm": 1.0196533203125,
"learning_rate": 5.219472072251154e-06,
"loss": 0.4598,
"step": 1090
},
{
"epoch": 0.6957350488444127,
"grad_norm": 1.0347778797149658,
"learning_rate": 5.12226479571868e-06,
"loss": 0.4458,
"step": 1095
},
{
"epoch": 0.6989119212135653,
"grad_norm": 1.0776883363723755,
"learning_rate": 5.025658610752568e-06,
"loss": 0.4322,
"step": 1100
},
{
"epoch": 0.7020887935827178,
"grad_norm": 1.0258809328079224,
"learning_rate": 4.929665422296532e-06,
"loss": 0.4376,
"step": 1105
},
{
"epoch": 0.7052656659518703,
"grad_norm": 1.0535441637039185,
"learning_rate": 4.834297059753682e-06,
"loss": 0.4496,
"step": 1110
},
{
"epoch": 0.708442538321023,
"grad_norm": 1.0777946710586548,
"learning_rate": 4.739565275528773e-06,
"loss": 0.455,
"step": 1115
},
{
"epoch": 0.7116194106901755,
"grad_norm": 1.0707253217697144,
"learning_rate": 4.645481743579949e-06,
"loss": 0.4795,
"step": 1120
},
{
"epoch": 0.7147962830593281,
"grad_norm": 0.963762640953064,
"learning_rate": 4.55205805798011e-06,
"loss": 0.4429,
"step": 1125
},
{
"epoch": 0.7179731554284806,
"grad_norm": 1.0962485074996948,
"learning_rate": 4.45930573148818e-06,
"loss": 0.494,
"step": 1130
},
{
"epoch": 0.7211500277976333,
"grad_norm": 1.0998930931091309,
"learning_rate": 4.367236194130375e-06,
"loss": 0.4401,
"step": 1135
},
{
"epoch": 0.7243269001667858,
"grad_norm": 0.9984025955200195,
"learning_rate": 4.275860791791638e-06,
"loss": 0.4433,
"step": 1140
},
{
"epoch": 0.7275037725359383,
"grad_norm": 1.0686578750610352,
"learning_rate": 4.185190784817478e-06,
"loss": 0.4553,
"step": 1145
},
{
"epoch": 0.730680644905091,
"grad_norm": 1.015057921409607,
"learning_rate": 4.095237346626345e-06,
"loss": 0.4366,
"step": 1150
},
{
"epoch": 0.7338575172742435,
"grad_norm": 1.0694704055786133,
"learning_rate": 4.006011562332702e-06,
"loss": 0.4503,
"step": 1155
},
{
"epoch": 0.7370343896433961,
"grad_norm": 1.125226616859436,
"learning_rate": 3.917524427380992e-06,
"loss": 0.4532,
"step": 1160
},
{
"epoch": 0.7402112620125486,
"grad_norm": 1.087633490562439,
"learning_rate": 3.829786846190648e-06,
"loss": 0.456,
"step": 1165
},
{
"epoch": 0.7433881343817013,
"grad_norm": 1.0903210639953613,
"learning_rate": 3.742809630812322e-06,
"loss": 0.4406,
"step": 1170
},
{
"epoch": 0.7465650067508538,
"grad_norm": 1.1029974222183228,
"learning_rate": 3.6566034995955e-06,
"loss": 0.4549,
"step": 1175
},
{
"epoch": 0.7497418791200063,
"grad_norm": 1.0449450016021729,
"learning_rate": 3.571179075867671e-06,
"loss": 0.4613,
"step": 1180
},
{
"epoch": 0.7529187514891589,
"grad_norm": 1.0384024381637573,
"learning_rate": 3.4865468866251794e-06,
"loss": 0.464,
"step": 1185
},
{
"epoch": 0.7560956238583115,
"grad_norm": 1.0278702974319458,
"learning_rate": 3.402717361235961e-06,
"loss": 0.4568,
"step": 1190
},
{
"epoch": 0.7592724962274641,
"grad_norm": 1.0673890113830566,
"learning_rate": 3.3197008301543497e-06,
"loss": 0.4594,
"step": 1195
},
{
"epoch": 0.7624493685966166,
"grad_norm": 1.0114554166793823,
"learning_rate": 3.2375075236480003e-06,
"loss": 0.4482,
"step": 1200
},
{
"epoch": 0.7656262409657691,
"grad_norm": 1.101015329360962,
"learning_rate": 3.156147570537209e-06,
"loss": 0.4596,
"step": 1205
},
{
"epoch": 0.7688031133349218,
"grad_norm": 1.1032803058624268,
"learning_rate": 3.075630996946729e-06,
"loss": 0.4442,
"step": 1210
},
{
"epoch": 0.7719799857040743,
"grad_norm": 1.0375980138778687,
"learning_rate": 2.9959677250702223e-06,
"loss": 0.4511,
"step": 1215
},
{
"epoch": 0.7751568580732269,
"grad_norm": 1.0073593854904175,
"learning_rate": 2.9171675719475355e-06,
"loss": 0.4384,
"step": 1220
},
{
"epoch": 0.7783337304423795,
"grad_norm": 1.045289158821106,
"learning_rate": 2.8392402482549397e-06,
"loss": 0.4719,
"step": 1225
},
{
"epoch": 0.7815106028115321,
"grad_norm": 0.984451949596405,
"learning_rate": 2.762195357108448e-06,
"loss": 0.4386,
"step": 1230
},
{
"epoch": 0.7846874751806846,
"grad_norm": 1.0313149690628052,
"learning_rate": 2.6860423928804135e-06,
"loss": 0.4381,
"step": 1235
},
{
"epoch": 0.7878643475498371,
"grad_norm": 1.0065529346466064,
"learning_rate": 2.6107907400295385e-06,
"loss": 0.467,
"step": 1240
},
{
"epoch": 0.7910412199189898,
"grad_norm": 1.0645593404769897,
"learning_rate": 2.53644967194439e-06,
"loss": 0.45,
"step": 1245
},
{
"epoch": 0.7942180922881423,
"grad_norm": 0.991398811340332,
"learning_rate": 2.4630283498006323e-06,
"loss": 0.4583,
"step": 1250
},
{
"epoch": 0.7973949646572949,
"grad_norm": 0.9970043897628784,
"learning_rate": 2.390535821432084e-06,
"loss": 0.4438,
"step": 1255
},
{
"epoch": 0.8005718370264474,
"grad_norm": 1.015313744544983,
"learning_rate": 2.3189810202157337e-06,
"loss": 0.4797,
"step": 1260
},
{
"epoch": 0.8037487093956001,
"grad_norm": 1.107275128364563,
"learning_rate": 2.2483727639708606e-06,
"loss": 0.4594,
"step": 1265
},
{
"epoch": 0.8069255817647526,
"grad_norm": 1.0633456707000732,
"learning_rate": 2.1787197538724147e-06,
"loss": 0.4462,
"step": 1270
},
{
"epoch": 0.8101024541339051,
"grad_norm": 1.0543006658554077,
"learning_rate": 2.1100305733787406e-06,
"loss": 0.4622,
"step": 1275
},
{
"epoch": 0.8132793265030578,
"grad_norm": 1.0409566164016724,
"learning_rate": 2.0423136871738227e-06,
"loss": 0.4397,
"step": 1280
},
{
"epoch": 0.8164561988722103,
"grad_norm": 1.0549768209457397,
"learning_rate": 1.9755774401241866e-06,
"loss": 0.4426,
"step": 1285
},
{
"epoch": 0.8196330712413629,
"grad_norm": 1.0367543697357178,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.4435,
"step": 1290
},
{
"epoch": 0.8228099436105154,
"grad_norm": 1.0637731552124023,
"learning_rate": 1.8450796377142566e-06,
"loss": 0.4147,
"step": 1295
},
{
"epoch": 0.825986815979668,
"grad_norm": 1.0210437774658203,
"learning_rate": 1.781334163819064e-06,
"loss": 0.479,
"step": 1300
},
{
"epoch": 0.8291636883488206,
"grad_norm": 1.0538122653961182,
"learning_rate": 1.718601490027606e-06,
"loss": 0.4366,
"step": 1305
},
{
"epoch": 0.8323405607179731,
"grad_norm": 1.0405868291854858,
"learning_rate": 1.6568893469934666e-06,
"loss": 0.4342,
"step": 1310
},
{
"epoch": 0.8355174330871257,
"grad_norm": 1.0490772724151611,
"learning_rate": 1.5962053396085075e-06,
"loss": 0.4526,
"step": 1315
},
{
"epoch": 0.8386943054562783,
"grad_norm": 1.0285438299179077,
"learning_rate": 1.5365569460656793e-06,
"loss": 0.4542,
"step": 1320
},
{
"epoch": 0.8418711778254309,
"grad_norm": 1.1192095279693604,
"learning_rate": 1.4779515169374914e-06,
"loss": 0.4391,
"step": 1325
},
{
"epoch": 0.8450480501945834,
"grad_norm": 0.988335371017456,
"learning_rate": 1.4203962742701893e-06,
"loss": 0.4377,
"step": 1330
},
{
"epoch": 0.848224922563736,
"grad_norm": 1.085839033126831,
"learning_rate": 1.3638983106937543e-06,
"loss": 0.4477,
"step": 1335
},
{
"epoch": 0.8514017949328886,
"grad_norm": 0.9356504678726196,
"learning_rate": 1.3084645885478797e-06,
"loss": 0.469,
"step": 1340
},
{
"epoch": 0.8545786673020411,
"grad_norm": 1.0390148162841797,
"learning_rate": 1.254101939023985e-06,
"loss": 0.4328,
"step": 1345
},
{
"epoch": 0.8577555396711937,
"grad_norm": 1.0623959302902222,
"learning_rate": 1.2008170613233971e-06,
"loss": 0.4422,
"step": 1350
},
{
"epoch": 0.8609324120403463,
"grad_norm": 1.1521815061569214,
"learning_rate": 1.1486165218317957e-06,
"loss": 0.4512,
"step": 1355
},
{
"epoch": 0.8641092844094989,
"grad_norm": 1.0439189672470093,
"learning_rate": 1.0975067533100337e-06,
"loss": 0.4304,
"step": 1360
},
{
"epoch": 0.8672861567786514,
"grad_norm": 0.9963186383247375,
"learning_rate": 1.0474940541014e-06,
"loss": 0.4328,
"step": 1365
},
{
"epoch": 0.8704630291478039,
"grad_norm": 1.0758447647094727,
"learning_rate": 9.9858458735548e-07,
"loss": 0.4542,
"step": 1370
},
{
"epoch": 0.8736399015169566,
"grad_norm": 1.106645941734314,
"learning_rate": 9.507843802686623e-07,
"loss": 0.448,
"step": 1375
},
{
"epoch": 0.8768167738861091,
"grad_norm": 1.107141137123108,
"learning_rate": 9.040993233413787e-07,
"loss": 0.4333,
"step": 1380
},
{
"epoch": 0.8799936462552617,
"grad_norm": 1.0433176755905151,
"learning_rate": 8.585351696522248e-07,
"loss": 0.4368,
"step": 1385
},
{
"epoch": 0.8831705186244143,
"grad_norm": 1.0009626150131226,
"learning_rate": 8.140975341489921e-07,
"loss": 0.4386,
"step": 1390
},
{
"epoch": 0.8863473909935669,
"grad_norm": 1.0583575963974,
"learning_rate": 7.707918929567282e-07,
"loss": 0.4414,
"step": 1395
},
{
"epoch": 0.8895242633627194,
"grad_norm": 1.043960452079773,
"learning_rate": 7.286235827029042e-07,
"loss": 0.4434,
"step": 1400
},
{
"epoch": 0.8927011357318719,
"grad_norm": 0.9781895875930786,
"learning_rate": 6.875977998597828e-07,
"loss": 0.4473,
"step": 1405
},
{
"epoch": 0.8958780081010246,
"grad_norm": 0.9682676792144775,
"learning_rate": 6.477196001040254e-07,
"loss": 0.4325,
"step": 1410
},
{
"epoch": 0.8990548804701771,
"grad_norm": 0.9957118630409241,
"learning_rate": 6.089938976936971e-07,
"loss": 0.415,
"step": 1415
},
{
"epoch": 0.9022317528393297,
"grad_norm": 1.020727276802063,
"learning_rate": 5.714254648626639e-07,
"loss": 0.4353,
"step": 1420
},
{
"epoch": 0.9054086252084822,
"grad_norm": 1.0520106554031372,
"learning_rate": 5.350189312324993e-07,
"loss": 0.4346,
"step": 1425
},
{
"epoch": 0.9085854975776348,
"grad_norm": 1.0706285238265991,
"learning_rate": 4.997787832419699e-07,
"loss": 0.4297,
"step": 1430
},
{
"epoch": 0.9117623699467874,
"grad_norm": 1.1210540533065796,
"learning_rate": 4.657093635941701e-07,
"loss": 0.4568,
"step": 1435
},
{
"epoch": 0.9149392423159399,
"grad_norm": 1.1412665843963623,
"learning_rate": 4.328148707213564e-07,
"loss": 0.454,
"step": 1440
},
{
"epoch": 0.9181161146850926,
"grad_norm": 1.0113401412963867,
"learning_rate": 4.010993582675693e-07,
"loss": 0.4604,
"step": 1445
},
{
"epoch": 0.9212929870542451,
"grad_norm": 0.9700157642364502,
"learning_rate": 3.7056673458909953e-07,
"loss": 0.4486,
"step": 1450
},
{
"epoch": 0.9244698594233977,
"grad_norm": 0.950957179069519,
"learning_rate": 3.412207622728458e-07,
"loss": 0.4198,
"step": 1455
},
{
"epoch": 0.9276467317925502,
"grad_norm": 0.957328200340271,
"learning_rate": 3.130650576726557e-07,
"loss": 0.4155,
"step": 1460
},
{
"epoch": 0.9308236041617028,
"grad_norm": 1.1100578308105469,
"learning_rate": 2.861030904636708e-07,
"loss": 0.4339,
"step": 1465
},
{
"epoch": 0.9340004765308554,
"grad_norm": 1.0033955574035645,
"learning_rate": 2.603381832147522e-07,
"loss": 0.4551,
"step": 1470
},
{
"epoch": 0.9371773489000079,
"grad_norm": 1.0719983577728271,
"learning_rate": 2.3577351097903157e-07,
"loss": 0.4368,
"step": 1475
},
{
"epoch": 0.9403542212691605,
"grad_norm": 1.0672398805618286,
"learning_rate": 2.1241210090265697e-07,
"loss": 0.4383,
"step": 1480
},
{
"epoch": 0.9435310936383131,
"grad_norm": 1.0421243906021118,
"learning_rate": 1.9025683185173727e-07,
"loss": 0.4509,
"step": 1485
},
{
"epoch": 0.9467079660074657,
"grad_norm": 1.0599042177200317,
"learning_rate": 1.6931043405758128e-07,
"loss": 0.438,
"step": 1490
},
{
"epoch": 0.9498848383766182,
"grad_norm": 0.9902300834655762,
"learning_rate": 1.4957548878025029e-07,
"loss": 0.4637,
"step": 1495
},
{
"epoch": 0.9530617107457708,
"grad_norm": 0.9928290843963623,
"learning_rate": 1.3105442799045576e-07,
"loss": 0.4422,
"step": 1500
},
{
"epoch": 0.9562385831149234,
"grad_norm": 1.024328589439392,
"learning_rate": 1.1374953406987244e-07,
"loss": 0.4593,
"step": 1505
},
{
"epoch": 0.9594154554840759,
"grad_norm": 0.9927188158035278,
"learning_rate": 9.766293952987449e-08,
"loss": 0.445,
"step": 1510
},
{
"epoch": 0.9625923278532285,
"grad_norm": 1.156173586845398,
"learning_rate": 8.279662674873679e-08,
"loss": 0.429,
"step": 1515
},
{
"epoch": 0.9657692002223811,
"grad_norm": 1.0015678405761719,
"learning_rate": 6.915242772734809e-08,
"loss": 0.4481,
"step": 1520
},
{
"epoch": 0.9689460725915336,
"grad_norm": 0.9944911599159241,
"learning_rate": 5.673202386345389e-08,
"loss": 0.4597,
"step": 1525
},
{
"epoch": 0.9721229449606862,
"grad_norm": 0.9489125609397888,
"learning_rate": 4.553694574444656e-08,
"loss": 0.4471,
"step": 1530
},
{
"epoch": 0.9752998173298387,
"grad_norm": 1.0264545679092407,
"learning_rate": 3.5568572958752935e-08,
"loss": 0.4408,
"step": 1535
},
{
"epoch": 0.9784766896989914,
"grad_norm": 1.0240846872329712,
"learning_rate": 2.682813392582917e-08,
"loss": 0.4575,
"step": 1540
},
{
"epoch": 0.9816535620681439,
"grad_norm": 0.995697021484375,
"learning_rate": 1.9316705744769626e-08,
"loss": 0.4261,
"step": 1545
},
{
"epoch": 0.9848304344372965,
"grad_norm": 1.057746171951294,
"learning_rate": 1.3035214061586365e-08,
"loss": 0.4561,
"step": 1550
},
{
"epoch": 0.9880073068064491,
"grad_norm": 1.13213312625885,
"learning_rate": 7.984432955133736e-09,
"loss": 0.4321,
"step": 1555
},
{
"epoch": 0.9911841791756016,
"grad_norm": 1.0002557039260864,
"learning_rate": 4.164984841715791e-09,
"loss": 0.4579,
"step": 1560
},
{
"epoch": 0.9943610515447542,
"grad_norm": 1.1303766965866089,
"learning_rate": 1.5773403983909697e-09,
"loss": 0.4427,
"step": 1565
},
{
"epoch": 0.9975379239139067,
"grad_norm": 1.0552936792373657,
"learning_rate": 2.218185049629451e-10,
"loss": 0.437,
"step": 1570
},
{
"epoch": 0.9994440473353983,
"step": 1573,
"total_flos": 4.7290819683968614e+17,
"train_loss": 0.5221994996298216,
"train_runtime": 21044.2981,
"train_samples_per_second": 1.795,
"train_steps_per_second": 0.075
}
],
"logging_steps": 5,
"max_steps": 1573,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.7290819683968614e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}