{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1860,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016148566814695196,
      "grad_norm": 0.5020289421081543,
      "learning_rate": 8.999999999999999e-05,
      "loss": 2.439466857910156,
      "step": 10
    },
    {
      "epoch": 0.03229713362939039,
      "grad_norm": 0.6233699321746826,
      "learning_rate": 0.00019,
      "loss": 2.4973506927490234,
      "step": 20
    },
    {
      "epoch": 0.04844570044408559,
      "grad_norm": 0.6460151076316833,
      "learning_rate": 0.00029,
      "loss": 2.579629898071289,
      "step": 30
    },
    {
      "epoch": 0.06459426725878079,
      "grad_norm": 0.5584802031517029,
      "learning_rate": 0.00039000000000000005,
      "loss": 2.5269956588745117,
      "step": 40
    },
    {
      "epoch": 0.08074283407347597,
      "grad_norm": 0.5267783403396606,
      "learning_rate": 0.00049,
      "loss": 2.5723337173461913,
      "step": 50
    },
    {
      "epoch": 0.09689140088817118,
      "grad_norm": 0.492374062538147,
      "learning_rate": 0.00059,
      "loss": 2.6084869384765623,
      "step": 60
    },
    {
      "epoch": 0.11303996770286637,
      "grad_norm": 0.533662736415863,
      "learning_rate": 0.00069,
      "loss": 2.527654838562012,
      "step": 70
    },
    {
      "epoch": 0.12918853451756157,
      "grad_norm": 0.568081796169281,
      "learning_rate": 0.00079,
      "loss": 2.584039497375488,
      "step": 80
    },
    {
      "epoch": 0.14533710133225677,
      "grad_norm": 0.5420770049095154,
      "learning_rate": 0.0008900000000000001,
      "loss": 2.5998212814331056,
      "step": 90
    },
    {
      "epoch": 0.16148566814695195,
      "grad_norm": 0.5972040295600891,
      "learning_rate": 0.00099,
      "loss": 2.61820011138916,
      "step": 100
    },
    {
      "epoch": 0.17763423496164715,
      "grad_norm": 0.622533917427063,
      "learning_rate": 0.0009999354806331361,
      "loss": 2.5819944381713866,
      "step": 110
    },
    {
      "epoch": 0.19378280177634236,
      "grad_norm": 0.6911935210227966,
      "learning_rate": 0.0009997124721002689,
      "loss": 2.5768102645874023,
      "step": 120
    },
    {
      "epoch": 0.20993136859103753,
      "grad_norm": 0.7095156908035278,
      "learning_rate": 0.000999330248902402,
      "loss": 2.6130306243896486,
      "step": 130
    },
    {
      "epoch": 0.22607993540573273,
      "grad_norm": 0.5791682600975037,
      "learning_rate": 0.0009987889328206437,
      "loss": 2.562555503845215,
      "step": 140
    },
    {
      "epoch": 0.24222850222042794,
      "grad_norm": 0.5518380403518677,
      "learning_rate": 0.0009980886963250907,
      "loss": 2.518760108947754,
      "step": 150
    },
    {
      "epoch": 0.25837706903512314,
      "grad_norm": 0.4897823631763458,
      "learning_rate": 0.000997229762519879,
      "loss": 2.6402866363525392,
      "step": 160
    },
    {
      "epoch": 0.2745256358498183,
      "grad_norm": 0.5050747394561768,
      "learning_rate": 0.0009962124050720978,
      "loss": 2.684323310852051,
      "step": 170
    },
    {
      "epoch": 0.29067420266451355,
      "grad_norm": 0.468423068523407,
      "learning_rate": 0.0009950369481245985,
      "loss": 2.6152185440063476,
      "step": 180
    },
    {
      "epoch": 0.3068227694792087,
      "grad_norm": 0.5091232657432556,
      "learning_rate": 0.0009937037661927161,
      "loss": 2.531853675842285,
      "step": 190
    },
    {
      "epoch": 0.3229713362939039,
      "grad_norm": 0.4922482967376709,
      "learning_rate": 0.0009922132840449458,
      "loss": 2.6094560623168945,
      "step": 200
    },
    {
      "epoch": 0.33911990310859913,
      "grad_norm": 0.5051465034484863,
      "learning_rate": 0.0009905659765676053,
      "loss": 2.559980583190918,
      "step": 210
    },
    {
      "epoch": 0.3552684699232943,
      "grad_norm": 0.4865105450153351,
      "learning_rate": 0.0009887623686135306,
      "loss": 2.508647346496582,
      "step": 220
    },
    {
      "epoch": 0.3714170367379895,
      "grad_norm": 0.5287356376647949,
      "learning_rate": 0.0009868030348348512,
      "loss": 2.6150222778320313,
      "step": 230
    },
    {
      "epoch": 0.3875656035526847,
      "grad_norm": 0.4643533229827881,
      "learning_rate": 0.0009846885994998983,
      "loss": 2.6150381088256838,
      "step": 240
    },
    {
      "epoch": 0.4037141703673799,
      "grad_norm": 0.46244189143180847,
      "learning_rate": 0.0009824197362943063,
      "loss": 2.5374935150146483,
      "step": 250
    },
    {
      "epoch": 0.41986273718207506,
      "grad_norm": 0.480276882648468,
      "learning_rate": 0.000979997168106366,
      "loss": 2.5654741287231446,
      "step": 260
    },
    {
      "epoch": 0.4360113039967703,
      "grad_norm": 0.5552269816398621,
      "learning_rate": 0.0009774216667967062,
      "loss": 2.6036794662475584,
      "step": 270
    },
    {
      "epoch": 0.45215987081146547,
      "grad_norm": 0.511289656162262,
      "learning_rate": 0.000974694052952366,
      "loss": 2.610031318664551,
      "step": 280
    },
    {
      "epoch": 0.4683084376261607,
      "grad_norm": 0.5037236213684082,
      "learning_rate": 0.000971815195625348,
      "loss": 2.516169548034668,
      "step": 290
    },
    {
      "epoch": 0.4844570044408559,
      "grad_norm": 0.5199276804924011,
      "learning_rate": 0.000968786012055726,
      "loss": 2.5069480895996095,
      "step": 300
    },
    {
      "epoch": 0.500605571255551,
      "grad_norm": 0.5875343680381775,
      "learning_rate": 0.0009656074673794017,
      "loss": 2.639841651916504,
      "step": 310
    },
    {
      "epoch": 0.5167541380702463,
      "grad_norm": 0.5489600896835327,
      "learning_rate": 0.0009622805743205998,
      "loss": 2.5628652572631836,
      "step": 320
    },
    {
      "epoch": 0.5329027048849415,
      "grad_norm": 0.4753468334674835,
      "learning_rate": 0.0009588063928692012,
      "loss": 2.5956233978271483,
      "step": 330
    },
    {
      "epoch": 0.5490512716996366,
      "grad_norm": 0.5152420997619629,
      "learning_rate": 0.0009551860299430173,
      "loss": 2.597700500488281,
      "step": 340
    },
    {
      "epoch": 0.5651998385143319,
      "grad_norm": 0.4520896375179291,
      "learning_rate": 0.0009514206390351116,
      "loss": 2.586415481567383,
      "step": 350
    },
    {
      "epoch": 0.5813484053290271,
      "grad_norm": 0.5123590230941772,
      "learning_rate": 0.0009475114198462837,
      "loss": 2.555033302307129,
      "step": 360
    },
    {
      "epoch": 0.5974969721437222,
      "grad_norm": 0.6008352637290955,
      "learning_rate": 0.0009434596179028271,
      "loss": 2.6199901580810545,
      "step": 370
    },
    {
      "epoch": 0.6136455389584174,
      "grad_norm": 0.4769132435321808,
      "learning_rate": 0.0009392665241596914,
      "loss": 2.5420787811279295,
      "step": 380
    },
    {
      "epoch": 0.6297941057731127,
      "grad_norm": 0.4474424421787262,
      "learning_rate": 0.0009349334745891666,
      "loss": 2.549270820617676,
      "step": 390
    },
    {
      "epoch": 0.6459426725878078,
      "grad_norm": 0.5046530365943909,
      "learning_rate": 0.0009304618497552253,
      "loss": 2.540346145629883,
      "step": 400
    },
    {
      "epoch": 0.662091239402503,
      "grad_norm": 0.5442773699760437,
      "learning_rate": 0.0009258530743736586,
      "loss": 2.550925636291504,
      "step": 410
    },
    {
      "epoch": 0.6782398062171983,
      "grad_norm": 0.45155641436576843,
      "learning_rate": 0.0009211086168581433,
      "loss": 2.5896928787231444,
      "step": 420
    },
    {
      "epoch": 0.6943883730318934,
      "grad_norm": 0.48448678851127625,
      "learning_rate": 0.0009162299888523867,
      "loss": 2.568522834777832,
      "step": 430
    },
    {
      "epoch": 0.7105369398465886,
      "grad_norm": 0.4634808897972107,
      "learning_rate": 0.0009112187447484979,
      "loss": 2.543706512451172,
      "step": 440
    },
    {
      "epoch": 0.7266855066612838,
      "grad_norm": 0.520962655544281,
      "learning_rate": 0.0009060764811917397,
      "loss": 2.4791580200195313,
      "step": 450
    },
    {
      "epoch": 0.742834073475979,
      "grad_norm": 0.495394766330719,
      "learning_rate": 0.0009008048365718167,
      "loss": 2.6086082458496094,
      "step": 460
    },
    {
      "epoch": 0.7589826402906742,
      "grad_norm": 0.4325544834136963,
      "learning_rate": 0.0008954054905008639,
      "loss": 2.5405605316162108,
      "step": 470
    },
    {
      "epoch": 0.7751312071053694,
      "grad_norm": 0.4992341697216034,
      "learning_rate": 0.0008898801632783013,
      "loss": 2.6021982192993165,
      "step": 480
    },
    {
      "epoch": 0.7912797739200645,
      "grad_norm": 0.5032821893692017,
      "learning_rate": 0.0008842306153427246,
      "loss": 2.5671119689941406,
      "step": 490
    },
    {
      "epoch": 0.8074283407347598,
      "grad_norm": 0.44175952672958374,
      "learning_rate": 0.000878458646711008,
      "loss": 2.5145410537719726,
      "step": 500
    },
    {
      "epoch": 0.823576907549455,
      "grad_norm": 0.48530301451683044,
      "learning_rate": 0.0008725660964047959,
      "loss": 2.4978832244873046,
      "step": 510
    },
    {
      "epoch": 0.8397254743641501,
      "grad_norm": 0.4604915976524353,
      "learning_rate": 0.0008665548418645672,
      "loss": 2.5596445083618162,
      "step": 520
    },
    {
      "epoch": 0.8558740411788454,
      "grad_norm": 0.45781826972961426,
      "learning_rate": 0.0008604267983514594,
      "loss": 2.6085268020629884,
      "step": 530
    },
    {
      "epoch": 0.8720226079935406,
      "grad_norm": 0.48819592595100403,
      "learning_rate": 0.000854183918337043,
      "loss": 2.5747554779052733,
      "step": 540
    },
    {
      "epoch": 0.8881711748082358,
      "grad_norm": 0.4929693043231964,
      "learning_rate": 0.0008478281908812387,
      "loss": 2.543058395385742,
      "step": 550
    },
    {
      "epoch": 0.9043197416229309,
      "grad_norm": 0.4913038909435272,
      "learning_rate": 0.0008413616409985779,
      "loss": 2.5399295806884767,
      "step": 560
    },
    {
      "epoch": 0.9204683084376262,
      "grad_norm": 0.47384563088417053,
      "learning_rate": 0.0008347863290130087,
      "loss": 2.5927974700927736,
      "step": 570
    },
    {
      "epoch": 0.9366168752523214,
      "grad_norm": 0.4775764048099518,
      "learning_rate": 0.0008281043499014498,
      "loss": 2.5593168258666994,
      "step": 580
    },
    {
      "epoch": 0.9527654420670165,
      "grad_norm": 0.5058761239051819,
      "learning_rate": 0.0008213178326263049,
      "loss": 2.552435111999512,
      "step": 590
    },
    {
      "epoch": 0.9689140088817118,
      "grad_norm": 0.4922596216201782,
      "learning_rate": 0.0008144289394571484,
      "loss": 2.5472679138183594,
      "step": 600
    },
    {
      "epoch": 0.985062575696407,
      "grad_norm": 0.49022358655929565,
      "learning_rate": 0.0008074398652817998,
      "loss": 2.5109460830688475,
      "step": 610
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.6876914501190186,
      "learning_rate": 0.0008003528369070043,
      "loss": 2.465944290161133,
      "step": 620
    },
    {
      "epoch": 1.0161485668146952,
      "grad_norm": 0.4119199812412262,
      "learning_rate": 0.0007931701123489439,
      "loss": 2.520769500732422,
      "step": 630
    },
    {
      "epoch": 1.0322971336293905,
      "grad_norm": 0.4147922992706299,
      "learning_rate": 0.000785893980113806,
      "loss": 2.5548782348632812,
      "step": 640
    },
    {
      "epoch": 1.0484457004440857,
      "grad_norm": 0.5313045382499695,
      "learning_rate": 0.0007785267584686366,
      "loss": 2.6331764221191407,
      "step": 650
    },
    {
      "epoch": 1.0645942672587807,
      "grad_norm": 0.45284605026245117,
      "learning_rate": 0.00077107079470271,
      "loss": 2.519462013244629,
      "step": 660
    },
    {
      "epoch": 1.080742834073476,
      "grad_norm": 0.5042719841003418,
      "learning_rate": 0.0007635284643796545,
      "loss": 2.4921422958374024,
      "step": 670
    },
    {
      "epoch": 1.0968914008881712,
      "grad_norm": 0.4403098225593567,
      "learning_rate": 0.0007559021705805671,
      "loss": 2.454839897155762,
      "step": 680
    },
    {
      "epoch": 1.1130399677028664,
      "grad_norm": 0.4963165819644928,
      "learning_rate": 0.0007481943431383622,
      "loss": 2.5821470260620116,
      "step": 690
    },
    {
      "epoch": 1.1291885345175616,
      "grad_norm": 0.4509197175502777,
      "learning_rate": 0.000740407437863596,
      "loss": 2.4973094940185545,
      "step": 700
    },
    {
      "epoch": 1.1453371013322569,
      "grad_norm": 0.49439796805381775,
      "learning_rate": 0.0007325439357620147,
      "loss": 2.6476114273071287,
      "step": 710
    },
    {
      "epoch": 1.1614856681469519,
      "grad_norm": 0.5243302583694458,
      "learning_rate": 0.0007246063422440747,
      "loss": 2.4758913040161135,
      "step": 720
    },
    {
      "epoch": 1.177634234961647,
      "grad_norm": 0.6043158769607544,
      "learning_rate": 0.0007165971863266878,
      "loss": 2.5820987701416014,
      "step": 730
    },
    {
      "epoch": 1.1937828017763423,
      "grad_norm": 0.44207850098609924,
      "learning_rate": 0.0007085190198274438,
      "loss": 2.4599235534667967,
      "step": 740
    },
    {
      "epoch": 1.2099313685910376,
      "grad_norm": 0.4254566431045532,
      "learning_rate": 0.0007003744165515704,
      "loss": 2.4942739486694334,
      "step": 750
    },
    {
      "epoch": 1.2260799354057328,
      "grad_norm": 0.41949278116226196,
      "learning_rate": 0.0006921659714718863,
      "loss": 2.505445098876953,
      "step": 760
    },
    {
      "epoch": 1.242228502220428,
      "grad_norm": 0.4123310148715973,
      "learning_rate": 0.0006838962999020094,
      "loss": 2.5693735122680663,
      "step": 770
    },
    {
      "epoch": 1.2583770690351233,
      "grad_norm": 0.4186009168624878,
      "learning_rate": 0.0006755680366630865,
      "loss": 2.4493398666381836,
      "step": 780
    },
    {
      "epoch": 1.2745256358498183,
      "grad_norm": 0.444654643535614,
      "learning_rate": 0.0006671838352443049,
      "loss": 2.5728691101074217,
      "step": 790
    },
    {
      "epoch": 1.2906742026645135,
      "grad_norm": 0.40202823281288147,
      "learning_rate": 0.0006587463669574584,
      "loss": 2.526685333251953,
      "step": 800
    },
    {
      "epoch": 1.3068227694792087,
      "grad_norm": 0.4873361885547638,
      "learning_rate": 0.0006502583200858335,
      "loss": 2.5635454177856447,
      "step": 810
    },
    {
      "epoch": 1.322971336293904,
      "grad_norm": 0.5394927859306335,
      "learning_rate": 0.0006417223990276883,
      "loss": 2.6018707275390627,
      "step": 820
    },
    {
      "epoch": 1.3391199031085992,
      "grad_norm": 0.5254472494125366,
      "learning_rate": 0.0006331413234345977,
      "loss": 2.5202842712402345,
      "step": 830
    },
    {
      "epoch": 1.3552684699232942,
      "grad_norm": 0.4611901342868805,
      "learning_rate": 0.0006245178273449383,
      "loss": 2.527310371398926,
      "step": 840
    },
    {
      "epoch": 1.3714170367379894,
      "grad_norm": 0.46280530095100403,
      "learning_rate": 0.0006158546583127886,
      "loss": 2.5010074615478515,
      "step": 850
    },
    {
      "epoch": 1.3875656035526847,
      "grad_norm": 0.47044530510902405,
      "learning_rate": 0.0006071545765325253,
      "loss": 2.658343505859375,
      "step": 860
    },
    {
      "epoch": 1.4037141703673799,
      "grad_norm": 0.5649057030677795,
      "learning_rate": 0.0005984203539593897,
      "loss": 2.514650344848633,
      "step": 870
    },
    {
      "epoch": 1.4198627371820751,
      "grad_norm": 0.5100232362747192,
      "learning_rate": 0.0005896547734263077,
      "loss": 2.4994720458984374,
      "step": 880
    },
    {
      "epoch": 1.4360113039967703,
      "grad_norm": 0.4990105926990509,
      "learning_rate": 0.0005808606277572453,
      "loss": 2.489163398742676,
      "step": 890
    },
    {
      "epoch": 1.4521598708114656,
      "grad_norm": 0.473297655582428,
      "learning_rate": 0.0005720407188773791,
      "loss": 2.534769630432129,
      "step": 900
    },
    {
      "epoch": 1.4683084376261606,
      "grad_norm": 0.5112643837928772,
      "learning_rate": 0.000563197856920368,
      "loss": 2.5247997283935546,
      "step": 910
    },
    {
      "epoch": 1.4844570044408558,
      "grad_norm": 0.4357326924800873,
      "learning_rate": 0.0005543348593330093,
      "loss": 2.508163642883301,
      "step": 920
    },
    {
      "epoch": 1.500605571255551,
      "grad_norm": 0.47014695405960083,
      "learning_rate": 0.0005454545499775651,
      "loss": 2.6127824783325195,
      "step": 930
    },
    {
      "epoch": 1.5167541380702463,
      "grad_norm": 0.4659437835216522,
      "learning_rate": 0.0005365597582320436,
      "loss": 2.4793100357055664,
      "step": 940
    },
    {
      "epoch": 1.5329027048849415,
      "grad_norm": 0.5408269166946411,
      "learning_rate": 0.0005276533180887248,
      "loss": 2.45506591796875,
      "step": 950
    },
    {
      "epoch": 1.5490512716996365,
      "grad_norm": 0.5870039463043213,
      "learning_rate": 0.000518738067251214,
      "loss": 2.480586814880371,
      "step": 960
    },
    {
      "epoch": 1.565199838514332,
      "grad_norm": 0.48131951689720154,
      "learning_rate": 0.0005098168462303141,
      "loss": 2.589830207824707,
      "step": 970
    },
    {
      "epoch": 1.581348405329027,
      "grad_norm": 0.4618188440799713,
      "learning_rate": 0.0005008924974390041,
      "loss": 2.5488056182861327,
      "step": 980
    },
    {
      "epoch": 1.5974969721437222,
      "grad_norm": 0.49082285165786743,
      "learning_rate": 0.0004919678642868092,
      "loss": 2.4781982421875,
      "step": 990
    },
    {
      "epoch": 1.6136455389584174,
      "grad_norm": 0.47744420170783997,
      "learning_rate": 0.0004830457902738558,
      "loss": 2.517325210571289,
      "step": 1000
    },
    {
      "epoch": 1.6297941057731127,
      "grad_norm": 0.507945716381073,
      "learning_rate": 0.0004741291180848961,
      "loss": 2.5076452255249024,
      "step": 1010
    },
    {
      "epoch": 1.645942672587808,
      "grad_norm": 0.42816001176834106,
      "learning_rate": 0.000465220688683594,
      "loss": 2.671817398071289,
      "step": 1020
    },
    {
      "epoch": 1.662091239402503,
      "grad_norm": 0.4258963167667389,
      "learning_rate": 0.00045632334040735764,
      "loss": 2.533784103393555,
      "step": 1030
    },
    {
      "epoch": 1.6782398062171984,
      "grad_norm": 0.494028240442276,
      "learning_rate": 0.00044743990806300917,
      "loss": 2.514291000366211,
      "step": 1040
    },
    {
      "epoch": 1.6943883730318934,
      "grad_norm": 0.4230322539806366,
      "learning_rate": 0.00043857322202358066,
      "loss": 2.5531696319580077,
      "step": 1050
    },
    {
      "epoch": 1.7105369398465886,
      "grad_norm": 0.5738111734390259,
      "learning_rate": 0.00042972610732652105,
      "loss": 2.5059112548828124,
      "step": 1060
    },
    {
      "epoch": 1.7266855066612838,
      "grad_norm": 0.5092839002609253,
      "learning_rate": 0.0004209013827736042,
      "loss": 2.5219795227050783,
      "step": 1070
    },
    {
      "epoch": 1.7428340734759789,
      "grad_norm": 0.4812857508659363,
      "learning_rate": 0.00041210186003282274,
      "loss": 2.5235408782958983,
      "step": 1080
    },
    {
      "epoch": 1.7589826402906743,
      "grad_norm": 0.47096627950668335,
      "learning_rate": 0.000403330342742556,
      "loss": 2.574551582336426,
      "step": 1090
    },
    {
      "epoch": 1.7751312071053693,
      "grad_norm": 0.4319113790988922,
      "learning_rate": 0.0003945896256182949,
      "loss": 2.610904502868652,
      "step": 1100
    },
    {
      "epoch": 1.7912797739200645,
      "grad_norm": 0.44047966599464417,
      "learning_rate": 0.0003858824935622115,
      "loss": 2.5323257446289062,
      "step": 1110
    },
    {
      "epoch": 1.8074283407347598,
      "grad_norm": 0.548047661781311,
      "learning_rate": 0.00037721172077585287,
      "loss": 2.5165468215942384,
      "step": 1120
    },
    {
      "epoch": 1.823576907549455,
      "grad_norm": 0.46526339650154114,
      "learning_rate": 0.00036858006987624723,
      "loss": 2.502303886413574,
      "step": 1130
    },
    {
      "epoch": 1.8397254743641502,
      "grad_norm": 0.4635223150253296,
      "learning_rate": 0.0003599902910156984,
      "loss": 2.5442089080810546,
      "step": 1140
    },
    {
      "epoch": 1.8558740411788452,
      "grad_norm": 0.5317935347557068,
      "learning_rate": 0.0003514451210055527,
      "loss": 2.600077247619629,
      "step": 1150
    },
    {
      "epoch": 1.8720226079935407,
      "grad_norm": 0.5463606119155884,
      "learning_rate": 0.00034294728244421756,
      "loss": 2.574476623535156,
      "step": 1160
    },
    {
      "epoch": 1.8881711748082357,
      "grad_norm": 0.4650241732597351,
      "learning_rate": 0.00033449948284970617,
      "loss": 2.500654411315918,
      "step": 1170
    },
    {
      "epoch": 1.904319741622931,
      "grad_norm": 0.40839987993240356,
      "learning_rate": 0.00032610441379698937,
      "loss": 2.4712839126586914,
      "step": 1180
    },
    {
      "epoch": 1.9204683084376262,
      "grad_norm": 0.49943020939826965,
      "learning_rate": 0.0003177647500604252,
      "loss": 2.5296091079711913,
      "step": 1190
    },
    {
      "epoch": 1.9366168752523214,
      "grad_norm": 0.4655015468597412,
      "learning_rate": 0.00030948314876154306,
      "loss": 2.5075130462646484,
      "step": 1200
    },
    {
      "epoch": 1.9527654420670166,
      "grad_norm": 0.45307499170303345,
      "learning_rate": 0.00030126224852245056,
      "loss": 2.464124298095703,
      "step": 1210
    },
    {
      "epoch": 1.9689140088817116,
      "grad_norm": 0.5530602931976318,
      "learning_rate": 0.0002931046686251365,
      "loss": 2.553061866760254,
      "step": 1220
    },
    {
      "epoch": 1.985062575696407,
      "grad_norm": 0.46495500206947327,
      "learning_rate": 0.0002850130081769334,
      "loss": 2.486197853088379,
      "step": 1230
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9545954465866089,
      "learning_rate": 0.00027698984528241036,
      "loss": 2.4371877670288087,
      "step": 1240
    },
    {
      "epoch": 2.016148566814695,
      "grad_norm": 0.48022302985191345,
      "learning_rate": 0.00026903773622195636,
      "loss": 2.512773895263672,
      "step": 1250
    },
    {
      "epoch": 2.0322971336293905,
      "grad_norm": 0.433242529630661,
      "learning_rate": 0.00026115921463731694,
      "loss": 2.5300994873046876,
      "step": 1260
    },
    {
      "epoch": 2.0484457004440855,
      "grad_norm": 0.47226834297180176,
      "learning_rate": 0.0002533567907243446,
      "loss": 2.421502113342285,
      "step": 1270
    },
    {
      "epoch": 2.064594267258781,
      "grad_norm": 0.5992064476013184,
      "learning_rate": 0.00024563295043321783,
      "loss": 2.5453359603881838,
      "step": 1280
    },
    {
      "epoch": 2.080742834073476,
      "grad_norm": 0.4261581599712372,
      "learning_rate": 0.0002379901546763879,
      "loss": 2.495037841796875,
      "step": 1290
    },
    {
      "epoch": 2.0968914008881714,
      "grad_norm": 0.4328082203865051,
      "learning_rate": 0.00023043083854449987,
      "loss": 2.464985466003418,
      "step": 1300
    },
    {
      "epoch": 2.1130399677028664,
      "grad_norm": 0.4510248303413391,
      "learning_rate": 0.00022295741053054296,
      "loss": 2.5308864593505858,
      "step": 1310
    },
    {
      "epoch": 2.1291885345175614,
      "grad_norm": 0.49293237924575806,
      "learning_rate": 0.00021557225176247353,
      "loss": 2.5278299331665037,
      "step": 1320
    },
    {
      "epoch": 2.145337101332257,
      "grad_norm": 0.4132377505302429,
      "learning_rate": 0.0002082777152445589,
      "loss": 2.398031234741211,
      "step": 1330
    },
    {
      "epoch": 2.161485668146952,
      "grad_norm": 0.5191354751586914,
      "learning_rate": 0.00020107612510768014,
      "loss": 2.5248828887939454,
      "step": 1340
    },
    {
      "epoch": 2.1776342349616473,
      "grad_norm": 0.5512005686759949,
      "learning_rate": 0.00019396977586883475,
      "loss": 2.4451154708862304,
      "step": 1350
    },
    {
      "epoch": 2.1937828017763423,
      "grad_norm": 0.42351678013801575,
      "learning_rate": 0.00018696093170007493,
      "loss": 2.573942756652832,
      "step": 1360
    },
    {
      "epoch": 2.2099313685910373,
      "grad_norm": 0.45575806498527527,
      "learning_rate": 0.00018005182570711366,
      "loss": 2.5537620544433595,
      "step": 1370
    },
    {
      "epoch": 2.226079935405733,
      "grad_norm": 0.46785682439804077,
      "learning_rate": 0.0001732446592178295,
      "loss": 2.491817092895508,
      "step": 1380
    },
    {
      "epoch": 2.242228502220428,
      "grad_norm": 0.45513054728507996,
      "learning_rate": 0.00016654160108089594,
      "loss": 2.5171764373779295,
      "step": 1390
    },
    {
      "epoch": 2.2583770690351233,
      "grad_norm": 0.43288710713386536,
      "learning_rate": 0.00015994478697475885,
      "loss": 2.5143251419067383,
      "step": 1400
    },
    {
      "epoch": 2.2745256358498183,
      "grad_norm": 0.4459301233291626,
      "learning_rate": 0.00015345631872718213,
      "loss": 2.5065849304199217,
      "step": 1410
    },
    {
      "epoch": 2.2906742026645137,
      "grad_norm": 0.501880407333374,
      "learning_rate": 0.00014707826364557985,
      "loss": 2.4833837509155274,
      "step": 1420
    },
    {
      "epoch": 2.3068227694792087,
      "grad_norm": 0.4559042453765869,
      "learning_rate": 0.00014081265385834557,
      "loss": 2.5120367050170898,
      "step": 1430
    },
    {
      "epoch": 2.3229713362939037,
      "grad_norm": 0.48048946261405945,
      "learning_rate": 0.000134661485667391,
      "loss": 2.4629817962646485,
      "step": 1440
    },
    {
      "epoch": 2.339119903108599,
      "grad_norm": 0.46413764357566833,
      "learning_rate": 0.0001286267189120986,
      "loss": 2.4572961807250975,
      "step": 1450
    },
    {
      "epoch": 2.355268469923294,
      "grad_norm": 0.4971129298210144,
      "learning_rate": 0.000122710276344893,
      "loss": 2.5448049545288085,
      "step": 1460
    },
    {
      "epoch": 2.3714170367379896,
      "grad_norm": 0.45018401741981506,
      "learning_rate": 0.00011691404301862746,
      "loss": 2.551463317871094,
      "step": 1470
    },
    {
      "epoch": 2.3875656035526847,
      "grad_norm": 0.49596303701400757,
      "learning_rate": 0.00011123986568598249,
      "loss": 2.5440711975097656,
      "step": 1480
    },
    {
      "epoch": 2.4037141703673797,
      "grad_norm": 0.47923141717910767,
      "learning_rate": 0.00010568955221106713,
      "loss": 2.45603084564209,
      "step": 1490
    },
    {
      "epoch": 2.419862737182075,
      "grad_norm": 0.4507387578487396,
      "learning_rate": 0.0001002648709934108,
      "loss": 2.458144187927246,
      "step": 1500
    },
    {
      "epoch": 2.43601130399677,
      "grad_norm": 0.46995073556900024,
      "learning_rate": 9.496755040452915e-05,
      "loss": 2.4886669158935546,
      "step": 1510
    },
    {
      "epoch": 2.4521598708114656,
      "grad_norm": 0.4773581326007843,
      "learning_rate": 8.979927823724321e-05,
      "loss": 2.548818016052246,
      "step": 1520
    },
    {
      "epoch": 2.4683084376261606,
      "grad_norm": 0.49100548028945923,
      "learning_rate": 8.476170116792736e-05,
      "loss": 2.5328250885009767,
      "step": 1530
    },
    {
      "epoch": 2.484457004440856,
      "grad_norm": 0.5164358615875244,
      "learning_rate": 7.985642423185718e-05,
      "loss": 2.4474281311035155,
      "step": 1540
    },
    {
      "epoch": 2.500605571255551,
      "grad_norm": 0.48450803756713867,
      "learning_rate": 7.508501031182585e-05,
      "loss": 2.470208168029785,
      "step": 1550
    },
    {
      "epoch": 2.5167541380702465,
      "grad_norm": 0.4779358208179474,
      "learning_rate": 7.044897964018949e-05,
      "loss": 2.443818283081055,
      "step": 1560
    },
    {
      "epoch": 2.5329027048849415,
      "grad_norm": 0.42919352650642395,
      "learning_rate": 6.594980931450223e-05,
      "loss": 2.4840776443481447,
      "step": 1570
    },
    {
      "epoch": 2.5490512716996365,
      "grad_norm": 0.4021783769130707,
      "learning_rate": 6.158893282689454e-05,
      "loss": 2.450935173034668,
      "step": 1580
    },
    {
      "epoch": 2.565199838514332,
      "grad_norm": 0.5504807829856873,
      "learning_rate": 5.7367739607344093e-05,
      "loss": 2.477644348144531,
      "step": 1590
    },
    {
      "epoch": 2.581348405329027,
      "grad_norm": 0.3885093331336975,
      "learning_rate": 5.328757458098665e-05,
      "loss": 2.449785041809082,
      "step": 1600
    },
    {
      "epoch": 2.5974969721437224,
      "grad_norm": 0.5313092470169067,
      "learning_rate": 4.934973773960572e-05,
      "loss": 2.564461898803711,
      "step": 1610
    },
    {
      "epoch": 2.6136455389584174,
      "grad_norm": 0.5081238150596619,
      "learning_rate": 4.5555483727438896e-05,
      "loss": 2.5144027709960937,
      "step": 1620
    },
    {
      "epoch": 2.6297941057731125,
      "grad_norm": 0.45351824164390564,
      "learning_rate": 4.190602144143207e-05,
      "loss": 2.5690656661987306,
      "step": 1630
    },
    {
      "epoch": 2.645942672587808,
      "grad_norm": 0.49749699234962463,
      "learning_rate": 3.840251364607045e-05,
      "loss": 2.5524566650390623,
      "step": 1640
    },
    {
      "epoch": 2.662091239402503,
      "grad_norm": 0.4736417233943939,
      "learning_rate": 3.50460766029066e-05,
      "loss": 2.437306022644043,
      "step": 1650
    },
    {
      "epoch": 2.6782398062171984,
      "grad_norm": 0.54044508934021,
      "learning_rate": 3.183777971490576e-05,
      "loss": 2.4329130172729494,
      "step": 1660
    },
    {
      "epoch": 2.6943883730318934,
      "grad_norm": 0.5381774306297302,
      "learning_rate": 2.8778645185720487e-05,
      "loss": 2.5384393692016602,
      "step": 1670
    },
    {
      "epoch": 2.7105369398465884,
      "grad_norm": 0.4624033570289612,
      "learning_rate": 2.5869647694003962e-05,
      "loss": 2.5221799850463866,
      "step": 1680
    },
    {
      "epoch": 2.726685506661284,
      "grad_norm": 0.49552807211875916,
      "learning_rate": 2.3111714082864887e-05,
      "loss": 2.519091987609863,
      "step": 1690
    },
    {
      "epoch": 2.742834073475979,
      "grad_norm": 0.4731680154800415,
      "learning_rate": 2.0505723064563886e-05,
      "loss": 2.4723621368408204,
      "step": 1700
    },
    {
      "epoch": 2.7589826402906743,
      "grad_norm": 0.44819337129592896,
      "learning_rate": 1.8052504940544613e-05,
      "loss": 2.5209144592285155,
      "step": 1710
    },
    {
      "epoch": 2.7751312071053693,
      "grad_norm": 0.4628264307975769,
      "learning_rate": 1.575284133688909e-05,
      "loss": 2.502878189086914,
      "step": 1720
    },
    {
      "epoch": 2.7912797739200643,
      "grad_norm": 0.4762296676635742,
      "learning_rate": 1.3607464955282257e-05,
      "loss": 2.4997129440307617,
      "step": 1730
    },
    {
      "epoch": 2.8074283407347598,
      "grad_norm": 0.4384547173976898,
      "learning_rate": 1.1617059339563806e-05,
      "loss": 2.6152523040771483,
      "step": 1740
    },
    {
      "epoch": 2.8235769075494552,
      "grad_norm": 0.4384756088256836,
      "learning_rate": 9.782258657942467e-06,
      "loss": 2.4833805084228517,
      "step": 1750
    },
    {
      "epoch": 2.8397254743641502,
      "grad_norm": 0.4431445300579071,
      "learning_rate": 8.103647500942112e-06,
      "loss": 2.500337028503418,
      "step": 1760
    },
    {
      "epoch": 2.8558740411788452,
      "grad_norm": 0.4854304790496826,
      "learning_rate": 6.581760695143934e-06,
      "loss": 2.4700483322143554,
      "step": 1770
    },
    {
      "epoch": 2.8720226079935407,
      "grad_norm": 0.4667441248893738,
      "learning_rate": 5.217083132783907e-06,
      "loss": 2.4867990493774412,
      "step": 1780
    },
    {
      "epoch": 2.8881711748082357,
      "grad_norm": 0.49439942836761475,
      "learning_rate": 4.010049617260203e-06,
      "loss": 2.515974426269531,
      "step": 1790
    },
    {
      "epoch": 2.904319741622931,
      "grad_norm": 0.43787845969200134,
      "learning_rate": 2.961044724599016e-06,
      "loss": 2.5289798736572267,
      "step": 1800
    },
    {
      "epoch": 2.920468308437626,
      "grad_norm": 0.4790705144405365,
      "learning_rate": 2.0704026809241215e-06,
      "loss": 2.4601634979248046,
      "step": 1810
    },
    {
      "epoch": 2.936616875252321,
      "grad_norm": 0.4470031261444092,
      "learning_rate": 1.338407255968288e-06,
      "loss": 2.4816938400268556,
      "step": 1820
    },
    {
      "epoch": 2.9527654420670166,
      "grad_norm": 0.4893916845321655,
      "learning_rate": 7.652916726604287e-07,
      "loss": 2.529372978210449,
      "step": 1830
    },
    {
      "epoch": 2.9689140088817116,
      "grad_norm": 0.431090384721756,
      "learning_rate": 3.5123853281793237e-07,
      "loss": 2.551029586791992,
      "step": 1840
    },
    {
      "epoch": 2.985062575696407,
      "grad_norm": 0.4225512742996216,
      "learning_rate": 9.637975896759077e-08,
      "loss": 2.5007448196411133,
      "step": 1850
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7053178548812866,
      "learning_rate": 7.965523131092667e-10,
      "loss": 2.677412414550781,
      "step": 1860
    }
  ],
  "logging_steps": 10,
  "max_steps": 1860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.131504576233472e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}