nothink_8b_100k / trainer_state.json
modrill's picture
Upload folder using huggingface_hub
038c25d verified
Raw
History Blame Contribute Delete
56 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 3126,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0064,
"grad_norm": 2.3469054747844766,
"learning_rate": 1.9148936170212767e-06,
"loss": 0.5239072322845459,
"step": 10
},
{
"epoch": 0.0128,
"grad_norm": 1.3731792313372932,
"learning_rate": 4.042553191489362e-06,
"loss": 0.4512838363647461,
"step": 20
},
{
"epoch": 0.0192,
"grad_norm": 1.2915003291066143,
"learning_rate": 6.170212765957447e-06,
"loss": 0.4259762763977051,
"step": 30
},
{
"epoch": 0.0256,
"grad_norm": 0.9428042712879645,
"learning_rate": 8.297872340425532e-06,
"loss": 0.3846761226654053,
"step": 40
},
{
"epoch": 0.032,
"grad_norm": 1.1140219220503393,
"learning_rate": 1.0425531914893619e-05,
"loss": 0.3693264961242676,
"step": 50
},
{
"epoch": 0.0384,
"grad_norm": 1.0868636121943922,
"learning_rate": 1.2553191489361702e-05,
"loss": 0.3520832538604736,
"step": 60
},
{
"epoch": 0.0448,
"grad_norm": 0.9655644749375284,
"learning_rate": 1.4680851063829789e-05,
"loss": 0.33546876907348633,
"step": 70
},
{
"epoch": 0.0512,
"grad_norm": 0.9875822564486919,
"learning_rate": 1.6808510638297873e-05,
"loss": 0.3456155300140381,
"step": 80
},
{
"epoch": 0.0576,
"grad_norm": 0.9445657536134299,
"learning_rate": 1.893617021276596e-05,
"loss": 0.32896125316619873,
"step": 90
},
{
"epoch": 0.064,
"grad_norm": 0.8741769235176083,
"learning_rate": 1.9999865800655937e-05,
"loss": 0.3304002285003662,
"step": 100
},
{
"epoch": 0.0704,
"grad_norm": 0.8158666037449533,
"learning_rate": 1.999879222751468e-05,
"loss": 0.32820682525634765,
"step": 110
},
{
"epoch": 0.0768,
"grad_norm": 0.7975671717413703,
"learning_rate": 1.9996645196489645e-05,
"loss": 0.3297333955764771,
"step": 120
},
{
"epoch": 0.0832,
"grad_norm": 0.7107586142224528,
"learning_rate": 1.9993424938083405e-05,
"loss": 0.3206096887588501,
"step": 130
},
{
"epoch": 0.0896,
"grad_norm": 0.7554089435348096,
"learning_rate": 1.99891317980189e-05,
"loss": 0.31907143592834475,
"step": 140
},
{
"epoch": 0.096,
"grad_norm": 0.7321972996848434,
"learning_rate": 1.9983766237202294e-05,
"loss": 0.3055716037750244,
"step": 150
},
{
"epoch": 0.1024,
"grad_norm": 0.7165826273420571,
"learning_rate": 1.997732883167352e-05,
"loss": 0.3128684997558594,
"step": 160
},
{
"epoch": 0.1088,
"grad_norm": 0.6919669508460716,
"learning_rate": 1.9969820272544416e-05,
"loss": 0.31496238708496094,
"step": 170
},
{
"epoch": 0.1152,
"grad_norm": 0.6464411891331754,
"learning_rate": 1.9961241365924545e-05,
"loss": 0.304630184173584,
"step": 180
},
{
"epoch": 0.1216,
"grad_norm": 0.8123267619504366,
"learning_rate": 1.9951593032834637e-05,
"loss": 0.3049804210662842,
"step": 190
},
{
"epoch": 0.128,
"grad_norm": 0.6735849444243486,
"learning_rate": 1.9940876309107722e-05,
"loss": 0.29981145858764646,
"step": 200
},
{
"epoch": 0.1344,
"grad_norm": 0.6733298901076364,
"learning_rate": 1.9929092345277917e-05,
"loss": 0.3078548669815063,
"step": 210
},
{
"epoch": 0.1408,
"grad_norm": 0.6258239601880233,
"learning_rate": 1.99162424064569e-05,
"loss": 0.3055974006652832,
"step": 220
},
{
"epoch": 0.1472,
"grad_norm": 0.6686785237738361,
"learning_rate": 1.9902327872198108e-05,
"loss": 0.29472782611846926,
"step": 230
},
{
"epoch": 0.1536,
"grad_norm": 0.622629401252819,
"learning_rate": 1.988735023634861e-05,
"loss": 0.28609018325805663,
"step": 240
},
{
"epoch": 0.16,
"grad_norm": 0.6106036320685414,
"learning_rate": 1.987131110688875e-05,
"loss": 0.2871884346008301,
"step": 250
},
{
"epoch": 0.1664,
"grad_norm": 0.5966642792994186,
"learning_rate": 1.9854212205759485e-05,
"loss": 0.2777689456939697,
"step": 260
},
{
"epoch": 0.1728,
"grad_norm": 0.6160635727815214,
"learning_rate": 1.9836055368677556e-05,
"loss": 0.28852453231811526,
"step": 270
},
{
"epoch": 0.1792,
"grad_norm": 0.6482595949889903,
"learning_rate": 1.9816842544938387e-05,
"loss": 0.28439571857452395,
"step": 280
},
{
"epoch": 0.1856,
"grad_norm": 0.5618125879312819,
"learning_rate": 1.9796575797206804e-05,
"loss": 0.28801932334899905,
"step": 290
},
{
"epoch": 0.192,
"grad_norm": 0.6692213222628999,
"learning_rate": 1.9775257301295616e-05,
"loss": 0.28852148056030275,
"step": 300
},
{
"epoch": 0.1984,
"grad_norm": 0.627289029944466,
"learning_rate": 1.9752889345931998e-05,
"loss": 0.2815143585205078,
"step": 310
},
{
"epoch": 0.2048,
"grad_norm": 0.5968195310508093,
"learning_rate": 1.9729474332511786e-05,
"loss": 0.27678637504577636,
"step": 320
},
{
"epoch": 0.2112,
"grad_norm": 0.5810704340065519,
"learning_rate": 1.9705014774841664e-05,
"loss": 0.28479642868041993,
"step": 330
},
{
"epoch": 0.2176,
"grad_norm": 0.7160577574731245,
"learning_rate": 1.9679513298869293e-05,
"loss": 0.28441829681396485,
"step": 340
},
{
"epoch": 0.224,
"grad_norm": 0.5530154626481694,
"learning_rate": 1.9652972642401378e-05,
"loss": 0.27650880813598633,
"step": 350
},
{
"epoch": 0.2304,
"grad_norm": 0.5411346112744626,
"learning_rate": 1.962539565480975e-05,
"loss": 0.2763069629669189,
"step": 360
},
{
"epoch": 0.2368,
"grad_norm": 0.6485613520321775,
"learning_rate": 1.959678529672546e-05,
"loss": 0.27900192737579343,
"step": 370
},
{
"epoch": 0.2432,
"grad_norm": 0.5642420099769282,
"learning_rate": 1.9567144639720937e-05,
"loss": 0.2747442960739136,
"step": 380
},
{
"epoch": 0.2496,
"grad_norm": 1.074554575011278,
"learning_rate": 1.95364768659802e-05,
"loss": 0.27370395660400393,
"step": 390
},
{
"epoch": 0.256,
"grad_norm": 0.5228549781390894,
"learning_rate": 1.9504785267957252e-05,
"loss": 0.2700518608093262,
"step": 400
},
{
"epoch": 0.2624,
"grad_norm": 0.5842380248151064,
"learning_rate": 1.9472073248022606e-05,
"loss": 0.278476881980896,
"step": 410
},
{
"epoch": 0.2688,
"grad_norm": 0.5446207083848129,
"learning_rate": 1.9438344318097982e-05,
"loss": 0.2713183879852295,
"step": 420
},
{
"epoch": 0.2752,
"grad_norm": 0.5529561315351936,
"learning_rate": 1.9403602099279306e-05,
"loss": 0.26999573707580565,
"step": 430
},
{
"epoch": 0.2816,
"grad_norm": 0.5134796198829531,
"learning_rate": 1.936785032144793e-05,
"loss": 0.2657029151916504,
"step": 440
},
{
"epoch": 0.288,
"grad_norm": 0.5546528361397897,
"learning_rate": 1.9331092822870204e-05,
"loss": 0.27097501754760744,
"step": 450
},
{
"epoch": 0.2944,
"grad_norm": 0.5277634405549436,
"learning_rate": 1.9293333549785414e-05,
"loss": 0.2600825309753418,
"step": 460
},
{
"epoch": 0.3008,
"grad_norm": 0.5564514235199555,
"learning_rate": 1.9254576555982097e-05,
"loss": 0.2580136775970459,
"step": 470
},
{
"epoch": 0.3072,
"grad_norm": 0.5401104213149478,
"learning_rate": 1.9214826002362855e-05,
"loss": 0.2737713813781738,
"step": 480
},
{
"epoch": 0.3136,
"grad_norm": 0.5349707104053464,
"learning_rate": 1.9174086156497626e-05,
"loss": 0.26891288757324217,
"step": 490
},
{
"epoch": 0.32,
"grad_norm": 0.4602795940666782,
"learning_rate": 1.9132361392165534e-05,
"loss": 0.26601734161376955,
"step": 500
},
{
"epoch": 0.3264,
"grad_norm": 0.5030400826450566,
"learning_rate": 1.9089656188885333e-05,
"loss": 0.26290915012359617,
"step": 510
},
{
"epoch": 0.3328,
"grad_norm": 0.5431367222435386,
"learning_rate": 1.904597513143447e-05,
"loss": 0.2582561016082764,
"step": 520
},
{
"epoch": 0.3392,
"grad_norm": 0.5404412264776388,
"learning_rate": 1.900132290935688e-05,
"loss": 0.27424280643463134,
"step": 530
},
{
"epoch": 0.3456,
"grad_norm": 0.5553400534109016,
"learning_rate": 1.8955704316459532e-05,
"loss": 0.2641714572906494,
"step": 540
},
{
"epoch": 0.352,
"grad_norm": 0.5845811733579757,
"learning_rate": 1.8909124250297762e-05,
"loss": 0.27608158588409426,
"step": 550
},
{
"epoch": 0.3584,
"grad_norm": 0.49797430077199206,
"learning_rate": 1.8861587711649466e-05,
"loss": 0.2599541902542114,
"step": 560
},
{
"epoch": 0.3648,
"grad_norm": 0.5174720409450502,
"learning_rate": 1.881309980397825e-05,
"loss": 0.25819385051727295,
"step": 570
},
{
"epoch": 0.3712,
"grad_norm": 0.5190565873161119,
"learning_rate": 1.876366573288551e-05,
"loss": 0.26655256748199463,
"step": 580
},
{
"epoch": 0.3776,
"grad_norm": 0.4279198538167247,
"learning_rate": 1.8713290805551564e-05,
"loss": 0.258132791519165,
"step": 590
},
{
"epoch": 0.384,
"grad_norm": 0.5064145194674196,
"learning_rate": 1.8661980430165886e-05,
"loss": 0.2618707180023193,
"step": 600
},
{
"epoch": 0.3904,
"grad_norm": 0.4536283905910432,
"learning_rate": 1.860974011534649e-05,
"loss": 0.2519304513931274,
"step": 610
},
{
"epoch": 0.3968,
"grad_norm": 0.4973309520583962,
"learning_rate": 1.855657546954853e-05,
"loss": 0.24944324493408204,
"step": 620
},
{
"epoch": 0.4032,
"grad_norm": 0.5500928080435353,
"learning_rate": 1.850249220046218e-05,
"loss": 0.2443866729736328,
"step": 630
},
{
"epoch": 0.4096,
"grad_norm": 0.49073137928584065,
"learning_rate": 1.8447496114399875e-05,
"loss": 0.24955644607543945,
"step": 640
},
{
"epoch": 0.416,
"grad_norm": 0.46639745568923063,
"learning_rate": 1.8391593115672933e-05,
"loss": 0.2560102462768555,
"step": 650
},
{
"epoch": 0.4224,
"grad_norm": 0.523648385153168,
"learning_rate": 1.833478920595769e-05,
"loss": 0.25697860717773435,
"step": 660
},
{
"epoch": 0.4288,
"grad_norm": 0.5149027856396282,
"learning_rate": 1.8277090483651165e-05,
"loss": 0.2461613655090332,
"step": 670
},
{
"epoch": 0.4352,
"grad_norm": 0.4755964135119324,
"learning_rate": 1.8218503143216348e-05,
"loss": 0.2509129762649536,
"step": 680
},
{
"epoch": 0.4416,
"grad_norm": 0.46745747049674496,
"learning_rate": 1.815903347451715e-05,
"loss": 0.2497704029083252,
"step": 690
},
{
"epoch": 0.448,
"grad_norm": 0.5174152385826454,
"learning_rate": 1.8098687862143162e-05,
"loss": 0.2417367458343506,
"step": 700
},
{
"epoch": 0.4544,
"grad_norm": 0.48272291743104295,
"learning_rate": 1.8037472784724192e-05,
"loss": 0.2532215118408203,
"step": 710
},
{
"epoch": 0.4608,
"grad_norm": 0.4758721891736801,
"learning_rate": 1.7975394814234725e-05,
"loss": 0.25108513832092283,
"step": 720
},
{
"epoch": 0.4672,
"grad_norm": 0.5679882500410512,
"learning_rate": 1.7912460615288384e-05,
"loss": 0.24127421379089356,
"step": 730
},
{
"epoch": 0.4736,
"grad_norm": 0.4750465930768755,
"learning_rate": 1.7848676944422402e-05,
"loss": 0.24612007141113282,
"step": 740
},
{
"epoch": 0.48,
"grad_norm": 0.5091469834068351,
"learning_rate": 1.7784050649372264e-05,
"loss": 0.24335813522338867,
"step": 750
},
{
"epoch": 0.4864,
"grad_norm": 0.469410500790412,
"learning_rate": 1.771858866833654e-05,
"loss": 0.2473893165588379,
"step": 760
},
{
"epoch": 0.4928,
"grad_norm": 0.49083256155940735,
"learning_rate": 1.765229802923201e-05,
"loss": 0.24752388000488282,
"step": 770
},
{
"epoch": 0.4992,
"grad_norm": 0.5706666842878829,
"learning_rate": 1.7585185848939142e-05,
"loss": 0.24637966156005858,
"step": 780
},
{
"epoch": 0.5056,
"grad_norm": 0.47579296659539233,
"learning_rate": 1.7517259332538054e-05,
"loss": 0.24853019714355468,
"step": 790
},
{
"epoch": 0.512,
"grad_norm": 0.463355613751907,
"learning_rate": 1.744852577253497e-05,
"loss": 0.2448970317840576,
"step": 800
},
{
"epoch": 0.5184,
"grad_norm": 0.4514002866865113,
"learning_rate": 1.73789925480793e-05,
"loss": 0.2453676462173462,
"step": 810
},
{
"epoch": 0.5248,
"grad_norm": 0.49508158173867095,
"learning_rate": 1.7308667124171454e-05,
"loss": 0.2437300443649292,
"step": 820
},
{
"epoch": 0.5312,
"grad_norm": 0.48745065244379576,
"learning_rate": 1.7237557050861376e-05,
"loss": 0.2651980400085449,
"step": 830
},
{
"epoch": 0.5376,
"grad_norm": 0.4742291487477817,
"learning_rate": 1.7165669962437994e-05,
"loss": 0.2370821237564087,
"step": 840
},
{
"epoch": 0.544,
"grad_norm": 0.4980927112459816,
"learning_rate": 1.7093013576609614e-05,
"loss": 0.24486489295959474,
"step": 850
},
{
"epoch": 0.5504,
"grad_norm": 0.4804729167489284,
"learning_rate": 1.701959569367535e-05,
"loss": 0.24334237575531006,
"step": 860
},
{
"epoch": 0.5568,
"grad_norm": 0.47157025675779163,
"learning_rate": 1.6945424195687702e-05,
"loss": 0.2422011375427246,
"step": 870
},
{
"epoch": 0.5632,
"grad_norm": 0.4629489129216441,
"learning_rate": 1.687050704560634e-05,
"loss": 0.24815070629119873,
"step": 880
},
{
"epoch": 0.5696,
"grad_norm": 0.45016959413292745,
"learning_rate": 1.6794852286443218e-05,
"loss": 0.2460242986679077,
"step": 890
},
{
"epoch": 0.576,
"grad_norm": 0.5156893082088817,
"learning_rate": 1.6718468040399072e-05,
"loss": 0.24848203659057616,
"step": 900
},
{
"epoch": 0.5824,
"grad_norm": 0.4378931208174623,
"learning_rate": 1.664136250799146e-05,
"loss": 0.246627140045166,
"step": 910
},
{
"epoch": 0.5888,
"grad_norm": 0.4415472676305731,
"learning_rate": 1.6563543967174328e-05,
"loss": 0.23817386627197265,
"step": 920
},
{
"epoch": 0.5952,
"grad_norm": 0.49430671895921824,
"learning_rate": 1.648502077244932e-05,
"loss": 0.2512770652770996,
"step": 930
},
{
"epoch": 0.6016,
"grad_norm": 0.4515039308185487,
"learning_rate": 1.6405801353968866e-05,
"loss": 0.24112675189971924,
"step": 940
},
{
"epoch": 0.608,
"grad_norm": 0.4627000727001904,
"learning_rate": 1.632589421663108e-05,
"loss": 0.24137527942657472,
"step": 950
},
{
"epoch": 0.6144,
"grad_norm": 0.44338620397115364,
"learning_rate": 1.6245307939166738e-05,
"loss": 0.24225597381591796,
"step": 960
},
{
"epoch": 0.6208,
"grad_norm": 0.47904796270467576,
"learning_rate": 1.616405117321825e-05,
"loss": 0.24234352111816407,
"step": 970
},
{
"epoch": 0.6272,
"grad_norm": 0.4538152649390663,
"learning_rate": 1.608213264241083e-05,
"loss": 0.23777763843536376,
"step": 980
},
{
"epoch": 0.6336,
"grad_norm": 0.44968988416126826,
"learning_rate": 1.599956114141595e-05,
"loss": 0.23575832843780517,
"step": 990
},
{
"epoch": 0.64,
"grad_norm": 0.42977872445886417,
"learning_rate": 1.5916345535007143e-05,
"loss": 0.24659445285797119,
"step": 1000
},
{
"epoch": 0.6464,
"grad_norm": 0.4331355661858506,
"learning_rate": 1.5832494757108296e-05,
"loss": 0.23725650310516358,
"step": 1010
},
{
"epoch": 0.6528,
"grad_norm": 0.4910169956013502,
"learning_rate": 1.5748017809834525e-05,
"loss": 0.23749151229858398,
"step": 1020
},
{
"epoch": 0.6592,
"grad_norm": 0.4561729908459038,
"learning_rate": 1.5662923762525705e-05,
"loss": 0.23774266242980957,
"step": 1030
},
{
"epoch": 0.6656,
"grad_norm": 0.412771632812648,
"learning_rate": 1.557722175077279e-05,
"loss": 0.22867307662963868,
"step": 1040
},
{
"epoch": 0.672,
"grad_norm": 0.4945279558616644,
"learning_rate": 1.549092097543707e-05,
"loss": 0.23382022380828857,
"step": 1050
},
{
"epoch": 0.6784,
"grad_norm": 0.43104138890098725,
"learning_rate": 1.540403070166232e-05,
"loss": 0.23120017051696778,
"step": 1060
},
{
"epoch": 0.6848,
"grad_norm": 0.4968228330148167,
"learning_rate": 1.531656025788014e-05,
"loss": 0.23886425495147706,
"step": 1070
},
{
"epoch": 0.6912,
"grad_norm": 0.45847060853886684,
"learning_rate": 1.5228519034808469e-05,
"loss": 0.23338027000427247,
"step": 1080
},
{
"epoch": 0.6976,
"grad_norm": 0.42793434571816935,
"learning_rate": 1.5139916484443392e-05,
"loss": 0.23655662536621094,
"step": 1090
},
{
"epoch": 0.704,
"grad_norm": 0.48678291680596986,
"learning_rate": 1.5050762119044388e-05,
"loss": 0.23051209449768068,
"step": 1100
},
{
"epoch": 0.7104,
"grad_norm": 0.44833198554008663,
"learning_rate": 1.4961065510113128e-05,
"loss": 0.23906197547912597,
"step": 1110
},
{
"epoch": 0.7168,
"grad_norm": 0.4234814123939215,
"learning_rate": 1.4870836287365861e-05,
"loss": 0.23278434276580812,
"step": 1120
},
{
"epoch": 0.7232,
"grad_norm": 0.4654905768837249,
"learning_rate": 1.4780084137699589e-05,
"loss": 0.23158683776855468,
"step": 1130
},
{
"epoch": 0.7296,
"grad_norm": 0.43961868752477146,
"learning_rate": 1.4688818804152106e-05,
"loss": 0.22675909996032714,
"step": 1140
},
{
"epoch": 0.736,
"grad_norm": 0.42562774558324196,
"learning_rate": 1.4597050084855985e-05,
"loss": 0.22868552207946777,
"step": 1150
},
{
"epoch": 0.7424,
"grad_norm": 0.45687064233512215,
"learning_rate": 1.4504787831986664e-05,
"loss": 0.23097996711730956,
"step": 1160
},
{
"epoch": 0.7488,
"grad_norm": 0.4635686742609441,
"learning_rate": 1.4412041950704734e-05,
"loss": 0.23512911796569824,
"step": 1170
},
{
"epoch": 0.7552,
"grad_norm": 0.4797333181243552,
"learning_rate": 1.4318822398092528e-05,
"loss": 0.2401136875152588,
"step": 1180
},
{
"epoch": 0.7616,
"grad_norm": 0.4617103914191676,
"learning_rate": 1.4225139182085145e-05,
"loss": 0.23787789344787597,
"step": 1190
},
{
"epoch": 0.768,
"grad_norm": 0.4250762715445169,
"learning_rate": 1.4131002360396008e-05,
"loss": 0.23605496883392335,
"step": 1200
},
{
"epoch": 0.7744,
"grad_norm": 0.42339605405743297,
"learning_rate": 1.4036422039437083e-05,
"loss": 0.2233750343322754,
"step": 1210
},
{
"epoch": 0.7808,
"grad_norm": 0.4299923683801405,
"learning_rate": 1.394140837323386e-05,
"loss": 0.23082523345947265,
"step": 1220
},
{
"epoch": 0.7872,
"grad_norm": 0.44423257008894196,
"learning_rate": 1.3845971562335241e-05,
"loss": 0.24129889011383057,
"step": 1230
},
{
"epoch": 0.7936,
"grad_norm": 0.3978033645755141,
"learning_rate": 1.3750121852718415e-05,
"loss": 0.2288198709487915,
"step": 1240
},
{
"epoch": 0.8,
"grad_norm": 0.4399126476249956,
"learning_rate": 1.3653869534688857e-05,
"loss": 0.22707161903381348,
"step": 1250
},
{
"epoch": 0.8064,
"grad_norm": 0.4149375868092431,
"learning_rate": 1.3557224941775583e-05,
"loss": 0.23408315181732178,
"step": 1260
},
{
"epoch": 0.8128,
"grad_norm": 0.4929393947595615,
"learning_rate": 1.3460198449621765e-05,
"loss": 0.22631232738494872,
"step": 1270
},
{
"epoch": 0.8192,
"grad_norm": 0.41111488215185005,
"learning_rate": 1.3362800474870774e-05,
"loss": 0.23379335403442383,
"step": 1280
},
{
"epoch": 0.8256,
"grad_norm": 0.47042814279371725,
"learning_rate": 1.3265041474047918e-05,
"loss": 0.23520944118499756,
"step": 1290
},
{
"epoch": 0.832,
"grad_norm": 0.4533891177739434,
"learning_rate": 1.31669319424378e-05,
"loss": 0.22470221519470215,
"step": 1300
},
{
"epoch": 0.8384,
"grad_norm": 0.44384875460645273,
"learning_rate": 1.306848241295757e-05,
"loss": 0.225510311126709,
"step": 1310
},
{
"epoch": 0.8448,
"grad_norm": 0.44320890790094497,
"learning_rate": 1.2969703455026139e-05,
"loss": 0.2236855983734131,
"step": 1320
},
{
"epoch": 0.8512,
"grad_norm": 0.4726942004750436,
"learning_rate": 1.2870605673429435e-05,
"loss": 0.22572519779205322,
"step": 1330
},
{
"epoch": 0.8576,
"grad_norm": 0.4372495947212519,
"learning_rate": 1.2771199707181898e-05,
"loss": 0.22141447067260742,
"step": 1340
},
{
"epoch": 0.864,
"grad_norm": 0.4728839955643096,
"learning_rate": 1.2671496228384292e-05,
"loss": 0.232117223739624,
"step": 1350
},
{
"epoch": 0.8704,
"grad_norm": 0.4622587752171778,
"learning_rate": 1.2571505941077962e-05,
"loss": 0.2343806028366089,
"step": 1360
},
{
"epoch": 0.8768,
"grad_norm": 0.4659963857947286,
"learning_rate": 1.2471239580095647e-05,
"loss": 0.22524487972259521,
"step": 1370
},
{
"epoch": 0.8832,
"grad_norm": 0.4171091606626652,
"learning_rate": 1.2370707909909024e-05,
"loss": 0.22701873779296874,
"step": 1380
},
{
"epoch": 0.8896,
"grad_norm": 0.4287777984112782,
"learning_rate": 1.2269921723473019e-05,
"loss": 0.2238243579864502,
"step": 1390
},
{
"epoch": 0.896,
"grad_norm": 0.3973506834352298,
"learning_rate": 1.216889184106713e-05,
"loss": 0.2237884521484375,
"step": 1400
},
{
"epoch": 0.9024,
"grad_norm": 0.5027278713122495,
"learning_rate": 1.2067629109133724e-05,
"loss": 0.22668781280517578,
"step": 1410
},
{
"epoch": 0.9088,
"grad_norm": 0.44727262522588773,
"learning_rate": 1.1966144399113626e-05,
"loss": 0.23116579055786132,
"step": 1420
},
{
"epoch": 0.9152,
"grad_norm": 0.4342706273296099,
"learning_rate": 1.1864448606278935e-05,
"loss": 0.23284392356872557,
"step": 1430
},
{
"epoch": 0.9216,
"grad_norm": 0.4383988403279945,
"learning_rate": 1.176255264856335e-05,
"loss": 0.23143448829650878,
"step": 1440
},
{
"epoch": 0.928,
"grad_norm": 0.46427441143257403,
"learning_rate": 1.166046746539002e-05,
"loss": 0.2259718656539917,
"step": 1450
},
{
"epoch": 0.9344,
"grad_norm": 0.44978980376372873,
"learning_rate": 1.1558204016497095e-05,
"loss": 0.22938432693481445,
"step": 1460
},
{
"epoch": 0.9408,
"grad_norm": 0.4437163119317134,
"learning_rate": 1.1455773280761123e-05,
"loss": 0.22493762969970704,
"step": 1470
},
{
"epoch": 0.9472,
"grad_norm": 0.44884690114425146,
"learning_rate": 1.135318625501835e-05,
"loss": 0.2168701171875,
"step": 1480
},
{
"epoch": 0.9536,
"grad_norm": 0.41444304379688296,
"learning_rate": 1.1250453952884125e-05,
"loss": 0.225518798828125,
"step": 1490
},
{
"epoch": 0.96,
"grad_norm": 0.4539288323461017,
"learning_rate": 1.1147587403570497e-05,
"loss": 0.22903010845184327,
"step": 1500
},
{
"epoch": 0.9664,
"grad_norm": 0.40236120878050685,
"learning_rate": 1.104459765070211e-05,
"loss": 0.22207205295562743,
"step": 1510
},
{
"epoch": 0.9728,
"grad_norm": 0.39967047986464427,
"learning_rate": 1.0941495751130601e-05,
"loss": 0.21291723251342773,
"step": 1520
},
{
"epoch": 0.9792,
"grad_norm": 0.426946829992971,
"learning_rate": 1.0838292773747532e-05,
"loss": 0.22067580223083497,
"step": 1530
},
{
"epoch": 0.9856,
"grad_norm": 0.4087278435213882,
"learning_rate": 1.0734999798296058e-05,
"loss": 0.22067804336547853,
"step": 1540
},
{
"epoch": 0.992,
"grad_norm": 0.5294633900184196,
"learning_rate": 1.0631627914181403e-05,
"loss": 0.21162500381469726,
"step": 1550
},
{
"epoch": 0.9984,
"grad_norm": 0.4415067076525386,
"learning_rate": 1.0528188219280339e-05,
"loss": 0.21919326782226561,
"step": 1560
},
{
"epoch": 1.00448,
"grad_norm": 0.41547053290884856,
"learning_rate": 1.0424691818749717e-05,
"loss": 0.18381383419036865,
"step": 1570
},
{
"epoch": 1.01088,
"grad_norm": 0.43321444402751025,
"learning_rate": 1.0321149823834227e-05,
"loss": 0.15648531913757324,
"step": 1580
},
{
"epoch": 1.01728,
"grad_norm": 0.3947322129353475,
"learning_rate": 1.0217573350673522e-05,
"loss": 0.15854500532150267,
"step": 1590
},
{
"epoch": 1.02368,
"grad_norm": 0.42607502946106895,
"learning_rate": 1.0113973519108797e-05,
"loss": 0.15481382608413696,
"step": 1600
},
{
"epoch": 1.03008,
"grad_norm": 0.4758549179319374,
"learning_rate": 1.0010361451488971e-05,
"loss": 0.16349189281463622,
"step": 1610
},
{
"epoch": 1.03648,
"grad_norm": 0.46557153733044215,
"learning_rate": 9.906748271476615e-06,
"loss": 0.15906081199645997,
"step": 1620
},
{
"epoch": 1.04288,
"grad_norm": 0.44269553099998354,
"learning_rate": 9.80314510285372e-06,
"loss": 0.1621649980545044,
"step": 1630
},
{
"epoch": 1.04928,
"grad_norm": 0.46102097780787,
"learning_rate": 9.69956306832748e-06,
"loss": 0.1558220386505127,
"step": 1640
},
{
"epoch": 1.05568,
"grad_norm": 0.4467135511309059,
"learning_rate": 9.596013288336137e-06,
"loss": 0.15889272689819336,
"step": 1650
},
{
"epoch": 1.06208,
"grad_norm": 0.4315325468630245,
"learning_rate": 9.492506879855143e-06,
"loss": 0.157787549495697,
"step": 1660
},
{
"epoch": 1.06848,
"grad_norm": 0.48718400293281816,
"learning_rate": 9.389054955203638e-06,
"loss": 0.15983572006225585,
"step": 1670
},
{
"epoch": 1.07488,
"grad_norm": 0.4443219575251595,
"learning_rate": 9.285668620851437e-06,
"loss": 0.16187721490859985,
"step": 1680
},
{
"epoch": 1.08128,
"grad_norm": 0.42156189066121047,
"learning_rate": 9.182358976226667e-06,
"loss": 0.1542805790901184,
"step": 1690
},
{
"epoch": 1.08768,
"grad_norm": 0.4405229533395611,
"learning_rate": 9.079137112524141e-06,
"loss": 0.1544118642807007,
"step": 1700
},
{
"epoch": 1.09408,
"grad_norm": 0.4392661868158838,
"learning_rate": 8.976014111514624e-06,
"loss": 0.152165949344635,
"step": 1710
},
{
"epoch": 1.10048,
"grad_norm": 0.4293858500547392,
"learning_rate": 8.873001044355103e-06,
"loss": 0.1639825701713562,
"step": 1720
},
{
"epoch": 1.10688,
"grad_norm": 0.42698960650358597,
"learning_rate": 8.770108970400206e-06,
"loss": 0.15804769992828369,
"step": 1730
},
{
"epoch": 1.11328,
"grad_norm": 0.4386828942931757,
"learning_rate": 8.667348936014874e-06,
"loss": 0.1597846269607544,
"step": 1740
},
{
"epoch": 1.11968,
"grad_norm": 0.4716490719728666,
"learning_rate": 8.56473197338845e-06,
"loss": 0.16282434463500978,
"step": 1750
},
{
"epoch": 1.12608,
"grad_norm": 0.480483144183952,
"learning_rate": 8.462269099350274e-06,
"loss": 0.15503289699554443,
"step": 1760
},
{
"epoch": 1.13248,
"grad_norm": 0.4321858297233622,
"learning_rate": 8.35997131418692e-06,
"loss": 0.16058000326156616,
"step": 1770
},
{
"epoch": 1.13888,
"grad_norm": 0.455044454802913,
"learning_rate": 8.25784960046123e-06,
"loss": 0.15212728977203369,
"step": 1780
},
{
"epoch": 1.14528,
"grad_norm": 0.48076978808751974,
"learning_rate": 8.155914921833242e-06,
"loss": 0.15667293071746827,
"step": 1790
},
{
"epoch": 1.15168,
"grad_norm": 0.4207881839362283,
"learning_rate": 8.054178221883121e-06,
"loss": 0.15629994869232178,
"step": 1800
},
{
"epoch": 1.15808,
"grad_norm": 0.4465287568017503,
"learning_rate": 7.952650422936303e-06,
"loss": 0.15517816543579102,
"step": 1810
},
{
"epoch": 1.16448,
"grad_norm": 0.47342713914577667,
"learning_rate": 7.851342424890868e-06,
"loss": 0.15276355743408204,
"step": 1820
},
{
"epoch": 1.17088,
"grad_norm": 0.46235414720257817,
"learning_rate": 7.750265104047347e-06,
"loss": 0.15710357427597046,
"step": 1830
},
{
"epoch": 1.17728,
"grad_norm": 0.4930916532423188,
"learning_rate": 7.649429311941053e-06,
"loss": 0.15528178215026855,
"step": 1840
},
{
"epoch": 1.18368,
"grad_norm": 0.4708847037489604,
"learning_rate": 7.548845874177066e-06,
"loss": 0.15652782917022706,
"step": 1850
},
{
"epoch": 1.19008,
"grad_norm": 0.42949278886690817,
"learning_rate": 7.4485255892680295e-06,
"loss": 0.1554331064224243,
"step": 1860
},
{
"epoch": 1.19648,
"grad_norm": 0.43399113540855677,
"learning_rate": 7.348479227474821e-06,
"loss": 0.15759174823760985,
"step": 1870
},
{
"epoch": 1.20288,
"grad_norm": 0.44907189406836595,
"learning_rate": 7.248717529650271e-06,
"loss": 0.15876730680465698,
"step": 1880
},
{
"epoch": 1.20928,
"grad_norm": 0.4366884973764217,
"learning_rate": 7.14925120608604e-06,
"loss": 0.15355956554412842,
"step": 1890
},
{
"epoch": 1.21568,
"grad_norm": 0.4154624081991762,
"learning_rate": 7.050090935362779e-06,
"loss": 0.1523200273513794,
"step": 1900
},
{
"epoch": 1.22208,
"grad_norm": 0.4204122167932454,
"learning_rate": 6.951247363203691e-06,
"loss": 0.15338459014892578,
"step": 1910
},
{
"epoch": 1.22848,
"grad_norm": 0.45325899705532907,
"learning_rate": 6.8527311013316e-06,
"loss": 0.1495049476623535,
"step": 1920
},
{
"epoch": 1.23488,
"grad_norm": 0.4274822961668358,
"learning_rate": 6.7545527263297215e-06,
"loss": 0.1549849510192871,
"step": 1930
},
{
"epoch": 1.24128,
"grad_norm": 0.43539874277708923,
"learning_rate": 6.656722778506151e-06,
"loss": 0.15497466325759887,
"step": 1940
},
{
"epoch": 1.24768,
"grad_norm": 0.4307968747341042,
"learning_rate": 6.559251760762275e-06,
"loss": 0.15383462905883788,
"step": 1950
},
{
"epoch": 1.25408,
"grad_norm": 0.43897948020096006,
"learning_rate": 6.462150137465194e-06,
"loss": 0.14782884120941162,
"step": 1960
},
{
"epoch": 1.26048,
"grad_norm": 0.4415989551440709,
"learning_rate": 6.3654283333242885e-06,
"loss": 0.157528293132782,
"step": 1970
},
{
"epoch": 1.26688,
"grad_norm": 0.4486937222750117,
"learning_rate": 6.269096732272009e-06,
"loss": 0.15642151832580567,
"step": 1980
},
{
"epoch": 1.27328,
"grad_norm": 0.43693098900388755,
"learning_rate": 6.173165676349103e-06,
"loss": 0.1492471933364868,
"step": 1990
},
{
"epoch": 1.27968,
"grad_norm": 0.4660736126625984,
"learning_rate": 6.077645464594291e-06,
"loss": 0.15759358406066895,
"step": 2000
},
{
"epoch": 1.2860800000000001,
"grad_norm": 0.40610381928333894,
"learning_rate": 5.982546351938572e-06,
"loss": 0.15414727926254274,
"step": 2010
},
{
"epoch": 1.29248,
"grad_norm": 0.4608452226379165,
"learning_rate": 5.887878548104271e-06,
"loss": 0.16232553720474244,
"step": 2020
},
{
"epoch": 1.29888,
"grad_norm": 0.40443189360675774,
"learning_rate": 5.793652216508929e-06,
"loss": 0.1495685577392578,
"step": 2030
},
{
"epoch": 1.30528,
"grad_norm": 0.4583075855809745,
"learning_rate": 5.699877473174185e-06,
"loss": 0.16189181804656982,
"step": 2040
},
{
"epoch": 1.31168,
"grad_norm": 0.4813490734723034,
"learning_rate": 5.606564385639723e-06,
"loss": 0.1553695797920227,
"step": 2050
},
{
"epoch": 1.31808,
"grad_norm": 0.43914460415648454,
"learning_rate": 5.513722971882434e-06,
"loss": 0.15075855255126952,
"step": 2060
},
{
"epoch": 1.3244799999999999,
"grad_norm": 0.4709002890713278,
"learning_rate": 5.421363199240897e-06,
"loss": 0.1548715353012085,
"step": 2070
},
{
"epoch": 1.33088,
"grad_norm": 0.44058954251166826,
"learning_rate": 5.329494983345311e-06,
"loss": 0.1570750117301941,
"step": 2080
},
{
"epoch": 1.33728,
"grad_norm": 0.44297821965041295,
"learning_rate": 5.238128187052938e-06,
"loss": 0.15714939832687377,
"step": 2090
},
{
"epoch": 1.34368,
"grad_norm": 0.43385750997981126,
"learning_rate": 5.147272619389267e-06,
"loss": 0.15720927715301514,
"step": 2100
},
{
"epoch": 1.35008,
"grad_norm": 0.40780705630732367,
"learning_rate": 5.0569380344949175e-06,
"loss": 0.15243306159973144,
"step": 2110
},
{
"epoch": 1.35648,
"grad_norm": 0.4440446022315159,
"learning_rate": 4.967134130578444e-06,
"loss": 0.14754390716552734,
"step": 2120
},
{
"epoch": 1.36288,
"grad_norm": 0.42892042333089464,
"learning_rate": 4.877870548875154e-06,
"loss": 0.1470635175704956,
"step": 2130
},
{
"epoch": 1.36928,
"grad_norm": 0.41542434327790073,
"learning_rate": 4.78915687261204e-06,
"loss": 0.14724797010421753,
"step": 2140
},
{
"epoch": 1.37568,
"grad_norm": 0.4519331341263826,
"learning_rate": 4.701002625978918e-06,
"loss": 0.14976963996887208,
"step": 2150
},
{
"epoch": 1.38208,
"grad_norm": 0.45911830958750144,
"learning_rate": 4.613417273105944e-06,
"loss": 0.14997029304504395,
"step": 2160
},
{
"epoch": 1.38848,
"grad_norm": 0.4364557617605061,
"learning_rate": 4.526410217047552e-06,
"loss": 0.15247538089752197,
"step": 2170
},
{
"epoch": 1.3948800000000001,
"grad_norm": 0.47321212205313934,
"learning_rate": 4.439990798772943e-06,
"loss": 0.15559486150741578,
"step": 2180
},
{
"epoch": 1.40128,
"grad_norm": 0.4392720177025314,
"learning_rate": 4.354168296163264e-06,
"loss": 0.15404411554336547,
"step": 2190
},
{
"epoch": 1.40768,
"grad_norm": 0.4172470540177006,
"learning_rate": 4.268951923015527e-06,
"loss": 0.14667506217956544,
"step": 2200
},
{
"epoch": 1.41408,
"grad_norm": 0.44298464733837534,
"learning_rate": 4.184350828053454e-06,
"loss": 0.15610729455947875,
"step": 2210
},
{
"epoch": 1.42048,
"grad_norm": 0.4235236468109531,
"learning_rate": 4.100374093945259e-06,
"loss": 0.15588223934173584,
"step": 2220
},
{
"epoch": 1.42688,
"grad_norm": 0.4363338589742784,
"learning_rate": 4.017030736328558e-06,
"loss": 0.1499912977218628,
"step": 2230
},
{
"epoch": 1.4332799999999999,
"grad_norm": 0.45754453052008726,
"learning_rate": 3.934329702842445e-06,
"loss": 0.14526530504226684,
"step": 2240
},
{
"epoch": 1.43968,
"grad_norm": 0.4069824746163237,
"learning_rate": 3.852279872166904e-06,
"loss": 0.14858238697052,
"step": 2250
},
{
"epoch": 1.44608,
"grad_norm": 0.4492016584721639,
"learning_rate": 3.770890053069597e-06,
"loss": 0.15298351049423217,
"step": 2260
},
{
"epoch": 1.45248,
"grad_norm": 0.42767171231247453,
"learning_rate": 3.6901689834601617e-06,
"loss": 0.14879769086837769,
"step": 2270
},
{
"epoch": 1.45888,
"grad_norm": 0.44002268528220406,
"learning_rate": 3.610125329452119e-06,
"loss": 0.14842283725738525,
"step": 2280
},
{
"epoch": 1.46528,
"grad_norm": 0.4246900218133806,
"learning_rate": 3.5307676844325e-06,
"loss": 0.14534516334533693,
"step": 2290
},
{
"epoch": 1.47168,
"grad_norm": 0.47963929577996556,
"learning_rate": 3.452104568139265e-06,
"loss": 0.15238602161407472,
"step": 2300
},
{
"epoch": 1.47808,
"grad_norm": 0.47240249550317165,
"learning_rate": 3.374144425746634e-06,
"loss": 0.15344419479370117,
"step": 2310
},
{
"epoch": 1.48448,
"grad_norm": 0.43660803164872536,
"learning_rate": 3.296895626958425e-06,
"loss": 0.14877289533615112,
"step": 2320
},
{
"epoch": 1.49088,
"grad_norm": 0.42023598853788297,
"learning_rate": 3.220366465109488e-06,
"loss": 0.1501172423362732,
"step": 2330
},
{
"epoch": 1.49728,
"grad_norm": 0.44716077077319133,
"learning_rate": 3.144565156275352e-06,
"loss": 0.15209732055664063,
"step": 2340
},
{
"epoch": 1.5036800000000001,
"grad_norm": 0.42433370562812966,
"learning_rate": 3.069499838390151e-06,
"loss": 0.15123218297958374,
"step": 2350
},
{
"epoch": 1.5100799999999999,
"grad_norm": 0.43923764707911134,
"learning_rate": 2.9951785703729487e-06,
"loss": 0.14909906387329103,
"step": 2360
},
{
"epoch": 1.51648,
"grad_norm": 0.4257285192860963,
"learning_rate": 2.921609331262538e-06,
"loss": 0.15349558591842652,
"step": 2370
},
{
"epoch": 1.52288,
"grad_norm": 0.4293950992523159,
"learning_rate": 2.8488000193608313e-06,
"loss": 0.14977505207061767,
"step": 2380
},
{
"epoch": 1.52928,
"grad_norm": 0.3876495666709967,
"learning_rate": 2.776758451384891e-06,
"loss": 0.15312135219573975,
"step": 2390
},
{
"epoch": 1.5356800000000002,
"grad_norm": 0.39448803369567287,
"learning_rate": 2.705492361627757e-06,
"loss": 0.14483506679534913,
"step": 2400
},
{
"epoch": 1.54208,
"grad_norm": 0.4482021383384413,
"learning_rate": 2.6350094011280903e-06,
"loss": 0.14908208847045898,
"step": 2410
},
{
"epoch": 1.54848,
"grad_norm": 0.4427224228196182,
"learning_rate": 2.5653171368487708e-06,
"loss": 0.14957340955734252,
"step": 2420
},
{
"epoch": 1.55488,
"grad_norm": 0.40812326693383305,
"learning_rate": 2.4964230508645148e-06,
"loss": 0.14661625623703003,
"step": 2430
},
{
"epoch": 1.56128,
"grad_norm": 0.4642610125321193,
"learning_rate": 2.4283345395586146e-06,
"loss": 0.14527816772460939,
"step": 2440
},
{
"epoch": 1.56768,
"grad_norm": 0.4681340493771961,
"learning_rate": 2.3610589128288596e-06,
"loss": 0.14932239055633545,
"step": 2450
},
{
"epoch": 1.57408,
"grad_norm": 0.44143167320183896,
"learning_rate": 2.2946033933027723e-06,
"loss": 0.14716734886169433,
"step": 2460
},
{
"epoch": 1.58048,
"grad_norm": 0.4378523330679974,
"learning_rate": 2.228975115562182e-06,
"loss": 0.14898219108581542,
"step": 2470
},
{
"epoch": 1.5868799999999998,
"grad_norm": 0.44843756895648756,
"learning_rate": 2.1641811253772727e-06,
"loss": 0.14232897758483887,
"step": 2480
},
{
"epoch": 1.59328,
"grad_norm": 0.47291812968473096,
"learning_rate": 2.1002283789501533e-06,
"loss": 0.14750297069549562,
"step": 2490
},
{
"epoch": 1.59968,
"grad_norm": 0.4645757234463273,
"learning_rate": 2.0371237421680546e-06,
"loss": 0.1520641803741455,
"step": 2500
},
{
"epoch": 1.60608,
"grad_norm": 0.48839894118009214,
"learning_rate": 1.9748739898662016e-06,
"loss": 0.15472906827926636,
"step": 2510
},
{
"epoch": 1.6124800000000001,
"grad_norm": 0.46325447015833676,
"learning_rate": 1.913485805100496e-06,
"loss": 0.15225306749343873,
"step": 2520
},
{
"epoch": 1.6188799999999999,
"grad_norm": 0.500660698075826,
"learning_rate": 1.852965778430017e-06,
"loss": 0.15347219705581666,
"step": 2530
},
{
"epoch": 1.62528,
"grad_norm": 0.4446655995893683,
"learning_rate": 1.7933204072094734e-06,
"loss": 0.14178776741027832,
"step": 2540
},
{
"epoch": 1.63168,
"grad_norm": 0.435805198787566,
"learning_rate": 1.7345560948916551e-06,
"loss": 0.14888372421264648,
"step": 2550
},
{
"epoch": 1.63808,
"grad_norm": 0.43419313537132415,
"learning_rate": 1.6766791503399638e-06,
"loss": 0.14740926027297974,
"step": 2560
},
{
"epoch": 1.6444800000000002,
"grad_norm": 0.446927783842065,
"learning_rate": 1.6196957871510988e-06,
"loss": 0.1472325325012207,
"step": 2570
},
{
"epoch": 1.65088,
"grad_norm": 0.4710136505286057,
"learning_rate": 1.5636121229879819e-06,
"loss": 0.15233616828918456,
"step": 2580
},
{
"epoch": 1.65728,
"grad_norm": 0.44426735021250846,
"learning_rate": 1.5084341789229673e-06,
"loss": 0.14666424989700316,
"step": 2590
},
{
"epoch": 1.66368,
"grad_norm": 0.45360688757314027,
"learning_rate": 1.454167878791426e-06,
"loss": 0.15017422437667846,
"step": 2600
},
{
"epoch": 1.67008,
"grad_norm": 0.43252749237002036,
"learning_rate": 1.4008190485557704e-06,
"loss": 0.13670945167541504,
"step": 2610
},
{
"epoch": 1.67648,
"grad_norm": 0.42876404381615957,
"learning_rate": 1.3483934156799938e-06,
"loss": 0.1442575454711914,
"step": 2620
},
{
"epoch": 1.68288,
"grad_norm": 0.46533715109502183,
"learning_rate": 1.2968966085147572e-06,
"loss": 0.14771933555603028,
"step": 2630
},
{
"epoch": 1.6892800000000001,
"grad_norm": 0.4779931036169498,
"learning_rate": 1.246334155693163e-06,
"loss": 0.15291142463684082,
"step": 2640
},
{
"epoch": 1.6956799999999999,
"grad_norm": 0.4765873696532621,
"learning_rate": 1.1967114855371876e-06,
"loss": 0.14684847593307496,
"step": 2650
},
{
"epoch": 1.70208,
"grad_norm": 0.4710421106584251,
"learning_rate": 1.148033925474913e-06,
"loss": 0.1526282548904419,
"step": 2660
},
{
"epoch": 1.70848,
"grad_norm": 0.46492240808726,
"learning_rate": 1.1003067014685798e-06,
"loss": 0.14880728721618652,
"step": 2670
},
{
"epoch": 1.71488,
"grad_norm": 0.4051940767391808,
"learning_rate": 1.0535349374535264e-06,
"loss": 0.15377696752548217,
"step": 2680
},
{
"epoch": 1.7212800000000001,
"grad_norm": 0.45252991663324355,
"learning_rate": 1.0077236547881031e-06,
"loss": 0.14457428455352783,
"step": 2690
},
{
"epoch": 1.7276799999999999,
"grad_norm": 0.42374467094591906,
"learning_rate": 9.628777717145765e-07,
"loss": 0.14815565347671508,
"step": 2700
},
{
"epoch": 1.73408,
"grad_norm": 0.45097627543056656,
"learning_rate": 9.190021028311158e-07,
"loss": 0.14766101837158202,
"step": 2710
},
{
"epoch": 1.74048,
"grad_norm": 0.4102704672516161,
"learning_rate": 8.761013585748979e-07,
"loss": 0.14567131996154786,
"step": 2720
},
{
"epoch": 1.74688,
"grad_norm": 0.4432145056669624,
"learning_rate": 8.341801447164111e-07,
"loss": 0.15420151948928834,
"step": 2730
},
{
"epoch": 1.75328,
"grad_norm": 0.4620617795493621,
"learning_rate": 7.932429618649795e-07,
"loss": 0.14718260765075683,
"step": 2740
},
{
"epoch": 1.75968,
"grad_norm": 0.3995145605741977,
"learning_rate": 7.532942049855785e-07,
"loss": 0.1427215576171875,
"step": 2750
},
{
"epoch": 1.76608,
"grad_norm": 0.4736529890517012,
"learning_rate": 7.143381629270074e-07,
"loss": 0.1467685580253601,
"step": 2760
},
{
"epoch": 1.77248,
"grad_norm": 0.4155673523575926,
"learning_rate": 6.763790179614371e-07,
"loss": 0.1465543270111084,
"step": 2770
},
{
"epoch": 1.77888,
"grad_norm": 0.4712162137928379,
"learning_rate": 6.394208453354067e-07,
"loss": 0.14595465660095214,
"step": 2780
},
{
"epoch": 1.78528,
"grad_norm": 0.4626916091428793,
"learning_rate": 6.034676128323124e-07,
"loss": 0.14494698047637938,
"step": 2790
},
{
"epoch": 1.79168,
"grad_norm": 0.4835096924945232,
"learning_rate": 5.685231803464275e-07,
"loss": 0.1401850938796997,
"step": 2800
},
{
"epoch": 1.7980800000000001,
"grad_norm": 0.46848803759410373,
"learning_rate": 5.345912994685143e-07,
"loss": 0.14994630813598633,
"step": 2810
},
{
"epoch": 1.8044799999999999,
"grad_norm": 0.45668819854752246,
"learning_rate": 5.016756130830547e-07,
"loss": 0.14242982864379883,
"step": 2820
},
{
"epoch": 1.81088,
"grad_norm": 0.4293096849122489,
"learning_rate": 4.697796549771538e-07,
"loss": 0.13955124616622924,
"step": 2830
},
{
"epoch": 1.81728,
"grad_norm": 0.4675817443378459,
"learning_rate": 4.389068494611615e-07,
"loss": 0.14673945903778077,
"step": 2840
},
{
"epoch": 1.82368,
"grad_norm": 0.46279606022530895,
"learning_rate": 4.0906051100103596e-07,
"loss": 0.13930708169937134,
"step": 2850
},
{
"epoch": 1.8300800000000002,
"grad_norm": 0.440471212876376,
"learning_rate": 3.8024384386251225e-07,
"loss": 0.1426056742668152,
"step": 2860
},
{
"epoch": 1.83648,
"grad_norm": 0.44280955239323283,
"learning_rate": 3.5245994176709264e-07,
"loss": 0.1487286329269409,
"step": 2870
},
{
"epoch": 1.84288,
"grad_norm": 0.4492207323139676,
"learning_rate": 3.257117875599125e-07,
"loss": 0.14853208065032958,
"step": 2880
},
{
"epoch": 1.84928,
"grad_norm": 0.4239142061654702,
"learning_rate": 3.00002252889503e-07,
"loss": 0.14546743631362916,
"step": 2890
},
{
"epoch": 1.85568,
"grad_norm": 0.47805467360149484,
"learning_rate": 2.753340978994945e-07,
"loss": 0.15466694831848143,
"step": 2900
},
{
"epoch": 1.86208,
"grad_norm": 0.44887677402906356,
"learning_rate": 2.5170997093229234e-07,
"loss": 0.14553663730621338,
"step": 2910
},
{
"epoch": 1.86848,
"grad_norm": 0.4306282565983831,
"learning_rate": 2.2913240824475103e-07,
"loss": 0.14442715644836426,
"step": 2920
},
{
"epoch": 1.87488,
"grad_norm": 0.41396020099122527,
"learning_rate": 2.07603833735891e-07,
"loss": 0.14591145515441895,
"step": 2930
},
{
"epoch": 1.8812799999999998,
"grad_norm": 0.45380623726387925,
"learning_rate": 1.871265586866633e-07,
"loss": 0.14740681648254395,
"step": 2940
},
{
"epoch": 1.88768,
"grad_norm": 0.4651060968137588,
"learning_rate": 1.6770278151182151e-07,
"loss": 0.15176613330841066,
"step": 2950
},
{
"epoch": 1.89408,
"grad_norm": 0.4268645571268607,
"learning_rate": 1.493345875238994e-07,
"loss": 0.14343844652175902,
"step": 2960
},
{
"epoch": 1.90048,
"grad_norm": 0.49261461241855176,
"learning_rate": 1.320239487093311e-07,
"loss": 0.1458448052406311,
"step": 2970
},
{
"epoch": 1.9068800000000001,
"grad_norm": 0.4832100615999477,
"learning_rate": 1.1577272351674607e-07,
"loss": 0.15465766191482544,
"step": 2980
},
{
"epoch": 1.9132799999999999,
"grad_norm": 0.4476533358866964,
"learning_rate": 1.0058265665744637e-07,
"loss": 0.14889904260635375,
"step": 2990
},
{
"epoch": 1.91968,
"grad_norm": 0.4567134572302328,
"learning_rate": 8.645537891809664e-08,
"loss": 0.1481461763381958,
"step": 3000
},
{
"epoch": 1.92608,
"grad_norm": 0.4366836886054917,
"learning_rate": 7.339240698564398e-08,
"loss": 0.14153001308441163,
"step": 3010
},
{
"epoch": 1.93248,
"grad_norm": 0.49956029767243015,
"learning_rate": 6.139514328449059e-08,
"loss": 0.15245096683502196,
"step": 3020
},
{
"epoch": 1.9388800000000002,
"grad_norm": 0.4524609352813635,
"learning_rate": 5.046487582592963e-08,
"loss": 0.14065054655075074,
"step": 3030
},
{
"epoch": 1.94528,
"grad_norm": 0.4509262586346184,
"learning_rate": 4.060277806986479e-08,
"loss": 0.14875115156173707,
"step": 3040
},
{
"epoch": 1.95168,
"grad_norm": 0.4321302618304759,
"learning_rate": 3.180990879883217e-08,
"loss": 0.14414600133895875,
"step": 3050
},
{
"epoch": 1.95808,
"grad_norm": 0.4383637710358457,
"learning_rate": 2.408721200432895e-08,
"loss": 0.15396742820739745,
"step": 3060
},
{
"epoch": 1.96448,
"grad_norm": 0.4463277669570913,
"learning_rate": 1.7435516785467842e-08,
"loss": 0.144897723197937,
"step": 3070
},
{
"epoch": 1.97088,
"grad_norm": 0.43022555013956015,
"learning_rate": 1.185553725996269e-08,
"loss": 0.13937076330184936,
"step": 3080
},
{
"epoch": 1.97728,
"grad_norm": 0.4573020672401508,
"learning_rate": 7.347872487470931e-09,
"loss": 0.14972403049468994,
"step": 3090
},
{
"epoch": 1.98368,
"grad_norm": 0.4526175505975703,
"learning_rate": 3.9130064052672616e-09,
"loss": 0.14847521781921386,
"step": 3100
},
{
"epoch": 1.9900799999999998,
"grad_norm": 0.4010602767640344,
"learning_rate": 1.5513077762985274e-09,
"loss": 0.14190553426742553,
"step": 3110
},
{
"epoch": 1.99648,
"grad_norm": 0.4690426965706053,
"learning_rate": 2.6303014958872773e-10,
"loss": 0.1439644455909729,
"step": 3120
},
{
"epoch": 2.0,
"step": 3126,
"total_flos": 508349128769536.0,
"train_loss": 0.205869834143156,
"train_runtime": 51800.7921,
"train_samples_per_second": 3.861,
"train_steps_per_second": 0.06
}
],
"logging_steps": 10,
"max_steps": 3126,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 508349128769536.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}