{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9990645463049579,
  "eval_steps": 500,
  "global_step": 534,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0018709073900841909,
      "grad_norm": 0.5139586586877397,
      "learning_rate": 1.8518518518518518e-07,
      "loss": 1.4813,
      "step": 1
    },
    {
      "epoch": 0.009354536950420954,
      "grad_norm": 0.5084917331083768,
      "learning_rate": 9.259259259259259e-07,
      "loss": 1.4869,
      "step": 5
    },
    {
      "epoch": 0.018709073900841908,
      "grad_norm": 0.5086882763520935,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 1.4912,
      "step": 10
    },
    {
      "epoch": 0.02806361085126286,
      "grad_norm": 0.49639126734512623,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 1.4813,
      "step": 15
    },
    {
      "epoch": 0.037418147801683815,
      "grad_norm": 0.5064168113795362,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 1.4849,
      "step": 20
    },
    {
      "epoch": 0.04677268475210477,
      "grad_norm": 0.474054321279664,
      "learning_rate": 4.62962962962963e-06,
      "loss": 1.4776,
      "step": 25
    },
    {
      "epoch": 0.05612722170252572,
      "grad_norm": 0.40667394555209146,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.4756,
      "step": 30
    },
    {
      "epoch": 0.06548175865294668,
      "grad_norm": 0.2821916256658075,
      "learning_rate": 6.481481481481482e-06,
      "loss": 1.4585,
      "step": 35
    },
    {
      "epoch": 0.07483629560336763,
      "grad_norm": 0.2369804283848295,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 1.4529,
      "step": 40
    },
    {
      "epoch": 0.08419083255378859,
      "grad_norm": 0.23752287185591942,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.4356,
      "step": 45
    },
    {
      "epoch": 0.09354536950420954,
      "grad_norm": 0.233703618586824,
      "learning_rate": 9.25925925925926e-06,
      "loss": 1.4195,
      "step": 50
    },
    {
      "epoch": 0.1028999064546305,
      "grad_norm": 0.21334448181657395,
      "learning_rate": 9.999892908320647e-06,
      "loss": 1.4009,
      "step": 55
    },
    {
      "epoch": 0.11225444340505145,
      "grad_norm": 0.21146314339445296,
      "learning_rate": 9.996145181203616e-06,
      "loss": 1.3904,
      "step": 60
    },
    {
      "epoch": 0.1216089803554724,
      "grad_norm": 0.20882987047024354,
      "learning_rate": 9.98704745668676e-06,
      "loss": 1.3718,
      "step": 65
    },
    {
      "epoch": 0.13096351730589337,
      "grad_norm": 0.21207151944716085,
      "learning_rate": 9.972609476841368e-06,
      "loss": 1.3591,
      "step": 70
    },
    {
      "epoch": 0.1403180542563143,
      "grad_norm": 0.22054679673319932,
      "learning_rate": 9.952846702217886e-06,
      "loss": 1.342,
      "step": 75
    },
    {
      "epoch": 0.14967259120673526,
      "grad_norm": 0.27057760894999794,
      "learning_rate": 9.92778029529039e-06,
      "loss": 1.3213,
      "step": 80
    },
    {
      "epoch": 0.15902712815715622,
      "grad_norm": 0.29764481463424675,
      "learning_rate": 9.897437097795257e-06,
      "loss": 1.2966,
      "step": 85
    },
    {
      "epoch": 0.16838166510757718,
      "grad_norm": 0.3056628343357791,
      "learning_rate": 9.861849601988384e-06,
      "loss": 1.2748,
      "step": 90
    },
    {
      "epoch": 0.17773620205799812,
      "grad_norm": 0.3099000095239784,
      "learning_rate": 9.821055915851647e-06,
      "loss": 1.2542,
      "step": 95
    },
    {
      "epoch": 0.18709073900841908,
      "grad_norm": 0.2668477842603494,
      "learning_rate": 9.775099722285934e-06,
      "loss": 1.2217,
      "step": 100
    },
    {
      "epoch": 0.19644527595884004,
      "grad_norm": 0.2595557434588695,
      "learning_rate": 9.72403023233439e-06,
      "loss": 1.2122,
      "step": 105
    },
    {
      "epoch": 0.205799812909261,
      "grad_norm": 0.2099058809011832,
      "learning_rate": 9.667902132486009e-06,
      "loss": 1.1948,
      "step": 110
    },
    {
      "epoch": 0.21515434985968196,
      "grad_norm": 0.20126316643616676,
      "learning_rate": 9.606775526115963e-06,
      "loss": 1.1774,
      "step": 115
    },
    {
      "epoch": 0.2245088868101029,
      "grad_norm": 0.17345905255518454,
      "learning_rate": 9.540715869125407e-06,
      "loss": 1.1681,
      "step": 120
    },
    {
      "epoch": 0.23386342376052385,
      "grad_norm": 0.25572107838360714,
      "learning_rate": 9.469793899849663e-06,
      "loss": 1.1716,
      "step": 125
    },
    {
      "epoch": 0.2432179607109448,
      "grad_norm": 0.15018874182675532,
      "learning_rate": 9.394085563309827e-06,
      "loss": 1.1569,
      "step": 130
    },
    {
      "epoch": 0.25257249766136575,
      "grad_norm": 0.15353642479045185,
      "learning_rate": 9.31367192988896e-06,
      "loss": 1.148,
      "step": 135
    },
    {
      "epoch": 0.26192703461178674,
      "grad_norm": 0.16296485857684445,
      "learning_rate": 9.228639108519867e-06,
      "loss": 1.1504,
      "step": 140
    },
    {
      "epoch": 0.27128157156220767,
      "grad_norm": 0.14482935533275312,
      "learning_rate": 9.139078154477512e-06,
      "loss": 1.1423,
      "step": 145
    },
    {
      "epoch": 0.2806361085126286,
      "grad_norm": 0.1566349158870611,
      "learning_rate": 9.045084971874738e-06,
      "loss": 1.137,
      "step": 150
    },
    {
      "epoch": 0.2899906454630496,
      "grad_norm": 0.14554164847096335,
      "learning_rate": 8.94676021096575e-06,
      "loss": 1.1363,
      "step": 155
    },
    {
      "epoch": 0.2993451824134705,
      "grad_norm": 0.1508011037252777,
      "learning_rate": 8.844209160367298e-06,
      "loss": 1.1304,
      "step": 160
    },
    {
      "epoch": 0.3086997193638915,
      "grad_norm": 0.14045125665102134,
      "learning_rate": 8.737541634312985e-06,
      "loss": 1.1332,
      "step": 165
    },
    {
      "epoch": 0.31805425631431244,
      "grad_norm": 0.14175186310915225,
      "learning_rate": 8.626871855061438e-06,
      "loss": 1.1341,
      "step": 170
    },
    {
      "epoch": 0.3274087932647334,
      "grad_norm": 0.14507855935792988,
      "learning_rate": 8.51231833058426e-06,
      "loss": 1.1302,
      "step": 175
    },
    {
      "epoch": 0.33676333021515437,
      "grad_norm": 0.1444068343697475,
      "learning_rate": 8.39400372766471e-06,
      "loss": 1.1333,
      "step": 180
    },
    {
      "epoch": 0.3461178671655753,
      "grad_norm": 0.14780446060701052,
      "learning_rate": 8.272054740543053e-06,
      "loss": 1.1324,
      "step": 185
    },
    {
      "epoch": 0.35547240411599623,
      "grad_norm": 0.14569418896265066,
      "learning_rate": 8.146601955249187e-06,
      "loss": 1.1173,
      "step": 190
    },
    {
      "epoch": 0.3648269410664172,
      "grad_norm": 0.14203531449915593,
      "learning_rate": 8.017779709767857e-06,
      "loss": 1.115,
      "step": 195
    },
    {
      "epoch": 0.37418147801683815,
      "grad_norm": 0.13692219577509893,
      "learning_rate": 7.88572595018617e-06,
      "loss": 1.1135,
      "step": 200
    },
    {
      "epoch": 0.38353601496725914,
      "grad_norm": 0.14078263930837534,
      "learning_rate": 7.750582082977468e-06,
      "loss": 1.1159,
      "step": 205
    },
    {
      "epoch": 0.3928905519176801,
      "grad_norm": 0.15520280646196877,
      "learning_rate": 7.612492823579744e-06,
      "loss": 1.1162,
      "step": 210
    },
    {
      "epoch": 0.402245088868101,
      "grad_norm": 0.14752002744879342,
      "learning_rate": 7.471606041430724e-06,
      "loss": 1.1131,
      "step": 215
    },
    {
      "epoch": 0.411599625818522,
      "grad_norm": 0.1509280644564414,
      "learning_rate": 7.328072601625558e-06,
      "loss": 1.1118,
      "step": 220
    },
    {
      "epoch": 0.42095416276894293,
      "grad_norm": 0.1616348908118278,
      "learning_rate": 7.18204620336671e-06,
      "loss": 1.1044,
      "step": 225
    },
    {
      "epoch": 0.4303086997193639,
      "grad_norm": 0.16062290714439983,
      "learning_rate": 7.033683215379002e-06,
      "loss": 1.1049,
      "step": 230
    },
    {
      "epoch": 0.43966323666978485,
      "grad_norm": 0.14900564691058804,
      "learning_rate": 6.883142508466054e-06,
      "loss": 1.1072,
      "step": 235
    },
    {
      "epoch": 0.4490177736202058,
      "grad_norm": 0.14583649841999052,
      "learning_rate": 6.730585285387465e-06,
      "loss": 1.1062,
      "step": 240
    },
    {
      "epoch": 0.4583723105706268,
      "grad_norm": 0.15656333466816236,
      "learning_rate": 6.57617490823885e-06,
      "loss": 1.1038,
      "step": 245
    },
    {
      "epoch": 0.4677268475210477,
      "grad_norm": 0.16642188631242563,
      "learning_rate": 6.420076723519615e-06,
      "loss": 1.112,
      "step": 250
    },
    {
      "epoch": 0.47708138447146864,
      "grad_norm": 0.266837548538648,
      "learning_rate": 6.26245788507579e-06,
      "loss": 1.1104,
      "step": 255
    },
    {
      "epoch": 0.4864359214218896,
      "grad_norm": 0.15891960204556085,
      "learning_rate": 6.103487175107508e-06,
      "loss": 1.1013,
      "step": 260
    },
    {
      "epoch": 0.49579045837231056,
      "grad_norm": 0.15199420768584646,
      "learning_rate": 5.943334823432777e-06,
      "loss": 1.0965,
      "step": 265
    },
    {
      "epoch": 0.5051449953227315,
      "grad_norm": 0.1649301263523591,
      "learning_rate": 5.782172325201155e-06,
      "loss": 1.1015,
      "step": 270
    },
    {
      "epoch": 0.5144995322731525,
      "grad_norm": 0.17068364714449247,
      "learning_rate": 5.620172257252427e-06,
      "loss": 1.1048,
      "step": 275
    },
    {
      "epoch": 0.5238540692235735,
      "grad_norm": 0.1560964872722369,
      "learning_rate": 5.457508093317013e-06,
      "loss": 1.1048,
      "step": 280
    },
    {
      "epoch": 0.5332086061739943,
      "grad_norm": 0.15004408686462048,
      "learning_rate": 5.294354018255945e-06,
      "loss": 1.1001,
      "step": 285
    },
    {
      "epoch": 0.5425631431244153,
      "grad_norm": 0.1563742845945151,
      "learning_rate": 5.130884741539367e-06,
      "loss": 1.0959,
      "step": 290
    },
    {
      "epoch": 0.5519176800748363,
      "grad_norm": 0.15549310467355393,
      "learning_rate": 4.967275310163241e-06,
      "loss": 1.097,
      "step": 295
    },
    {
      "epoch": 0.5612722170252572,
      "grad_norm": 0.17033773254724513,
      "learning_rate": 4.803700921204659e-06,
      "loss": 1.0991,
      "step": 300
    },
    {
      "epoch": 0.5706267539756782,
      "grad_norm": 0.16793233585056505,
      "learning_rate": 4.640336734216403e-06,
      "loss": 1.1011,
      "step": 305
    },
    {
      "epoch": 0.5799812909260992,
      "grad_norm": 0.15858153553414958,
      "learning_rate": 4.477357683661734e-06,
      "loss": 1.0953,
      "step": 310
    },
    {
      "epoch": 0.5893358278765201,
      "grad_norm": 0.18459161802082374,
      "learning_rate": 4.314938291590161e-06,
      "loss": 1.0905,
      "step": 315
    },
    {
      "epoch": 0.598690364826941,
      "grad_norm": 0.16567092468891204,
      "learning_rate": 4.1532524807548776e-06,
      "loss": 1.0906,
      "step": 320
    },
    {
      "epoch": 0.608044901777362,
      "grad_norm": 0.16035972271379115,
      "learning_rate": 3.992473388371914e-06,
      "loss": 1.0924,
      "step": 325
    },
    {
      "epoch": 0.617399438727783,
      "grad_norm": 0.18108290101376773,
      "learning_rate": 3.832773180720475e-06,
      "loss": 1.092,
      "step": 330
    },
    {
      "epoch": 0.6267539756782039,
      "grad_norm": 0.16657617125834437,
      "learning_rate": 3.6743228687829596e-06,
      "loss": 1.0887,
      "step": 335
    },
    {
      "epoch": 0.6361085126286249,
      "grad_norm": 0.1783833220121851,
      "learning_rate": 3.517292125122146e-06,
      "loss": 1.0938,
      "step": 340
    },
    {
      "epoch": 0.6454630495790459,
      "grad_norm": 0.15395137633162537,
      "learning_rate": 3.3618491021915334e-06,
      "loss": 1.0902,
      "step": 345
    },
    {
      "epoch": 0.6548175865294668,
      "grad_norm": 0.16256299187537024,
      "learning_rate": 3.2081602522734987e-06,
      "loss": 1.0875,
      "step": 350
    },
    {
      "epoch": 0.6641721234798877,
      "grad_norm": 0.15781917797031997,
      "learning_rate": 3.056390149238022e-06,
      "loss": 1.0897,
      "step": 355
    },
    {
      "epoch": 0.6735266604303087,
      "grad_norm": 0.15450725652248593,
      "learning_rate": 2.906701312312861e-06,
      "loss": 1.0883,
      "step": 360
    },
    {
      "epoch": 0.6828811973807296,
      "grad_norm": 0.16092272691343973,
      "learning_rate": 2.759254032053888e-06,
      "loss": 1.0908,
      "step": 365
    },
    {
      "epoch": 0.6922357343311506,
      "grad_norm": 0.15881876352311045,
      "learning_rate": 2.614206198701958e-06,
      "loss": 1.0964,
      "step": 370
    },
    {
      "epoch": 0.7015902712815716,
      "grad_norm": 0.16123816798728421,
      "learning_rate": 2.471713133110078e-06,
      "loss": 1.0826,
      "step": 375
    },
    {
      "epoch": 0.7109448082319925,
      "grad_norm": 0.17326937796971717,
      "learning_rate": 2.3319274204219427e-06,
      "loss": 1.0857,
      "step": 380
    },
    {
      "epoch": 0.7202993451824135,
      "grad_norm": 0.15824306754559836,
      "learning_rate": 2.1949987466799524e-06,
      "loss": 1.0926,
      "step": 385
    },
    {
      "epoch": 0.7296538821328344,
      "grad_norm": 0.15676499958123513,
      "learning_rate": 2.061073738537635e-06,
      "loss": 1.091,
      "step": 390
    },
    {
      "epoch": 0.7390084190832554,
      "grad_norm": 0.15899394910820866,
      "learning_rate": 1.9302958062481673e-06,
      "loss": 1.0845,
      "step": 395
    },
    {
      "epoch": 0.7483629560336763,
      "grad_norm": 0.16579475502144433,
      "learning_rate": 1.8028049900970768e-06,
      "loss": 1.0854,
      "step": 400
    },
    {
      "epoch": 0.7577174929840973,
      "grad_norm": 0.15812690411171398,
      "learning_rate": 1.6787378104435931e-06,
      "loss": 1.0867,
      "step": 405
    },
    {
      "epoch": 0.7670720299345183,
      "grad_norm": 0.1562749678781465,
      "learning_rate": 1.5582271215312294e-06,
      "loss": 1.0869,
      "step": 410
    },
    {
      "epoch": 0.7764265668849392,
      "grad_norm": 0.1573904723461426,
      "learning_rate": 1.4414019692241437e-06,
      "loss": 1.0968,
      "step": 415
    },
    {
      "epoch": 0.7857811038353602,
      "grad_norm": 0.1663255000536237,
      "learning_rate": 1.3283874528215735e-06,
      "loss": 1.0896,
      "step": 420
    },
    {
      "epoch": 0.7951356407857811,
      "grad_norm": 0.1565901532094871,
      "learning_rate": 1.2193045910983864e-06,
      "loss": 1.0892,
      "step": 425
    },
    {
      "epoch": 0.804490177736202,
      "grad_norm": 0.15886041285660657,
      "learning_rate": 1.1142701927151456e-06,
      "loss": 1.086,
      "step": 430
    },
    {
      "epoch": 0.813844714686623,
      "grad_norm": 0.1653805801993096,
      "learning_rate": 1.013396731136465e-06,
      "loss": 1.0846,
      "step": 435
    },
    {
      "epoch": 0.823199251637044,
      "grad_norm": 0.1659489588893349,
      "learning_rate": 9.167922241916055e-07,
      "loss": 1.0882,
      "step": 440
    },
    {
      "epoch": 0.8325537885874649,
      "grad_norm": 0.14935331942375346,
      "learning_rate": 8.245601184062851e-07,
      "loss": 1.0919,
      "step": 445
    },
    {
      "epoch": 0.8419083255378859,
      "grad_norm": 0.1612238398597313,
      "learning_rate": 7.367991782295392e-07,
      "loss": 1.0956,
      "step": 450
    },
    {
      "epoch": 0.8512628624883068,
      "grad_norm": 0.16096449999893725,
      "learning_rate": 6.536033802742814e-07,
      "loss": 1.0892,
      "step": 455
    },
    {
      "epoch": 0.8606173994387278,
      "grad_norm": 0.18591172254707147,
      "learning_rate": 5.750618126847912e-07,
      "loss": 1.0814,
      "step": 460
    },
    {
      "epoch": 0.8699719363891487,
      "grad_norm": 0.16230740670761631,
      "learning_rate": 5.012585797388936e-07,
      "loss": 1.0863,
      "step": 465
    },
    {
      "epoch": 0.8793264733395697,
      "grad_norm": 0.16399177374939766,
      "learning_rate": 4.322727117869951e-07,
      "loss": 1.0836,
      "step": 470
    },
    {
      "epoch": 0.8886810102899907,
      "grad_norm": 0.19550575898363698,
      "learning_rate": 3.6817808062440953e-07,
      "loss": 1.091,
      "step": 475
    },
    {
      "epoch": 0.8980355472404116,
      "grad_norm": 0.15359012389254986,
      "learning_rate": 3.0904332038757977e-07,
      "loss": 1.0937,
      "step": 480
    },
    {
      "epoch": 0.9073900841908326,
      "grad_norm": 0.1606805442454799,
      "learning_rate": 2.5493175405893076e-07,
      "loss": 1.0848,
      "step": 485
    },
    {
      "epoch": 0.9167446211412535,
      "grad_norm": 0.16036880753390018,
      "learning_rate": 2.0590132565903475e-07,
      "loss": 1.0904,
      "step": 490
    },
    {
      "epoch": 0.9260991580916744,
      "grad_norm": 0.16046720313774515,
      "learning_rate": 1.6200453819870122e-07,
      "loss": 1.0839,
      "step": 495
    },
    {
      "epoch": 0.9354536950420954,
      "grad_norm": 0.16013022562616708,
      "learning_rate": 1.232883974574367e-07,
      "loss": 1.0876,
      "step": 500
    },
    {
      "epoch": 0.9448082319925164,
      "grad_norm": 0.1680369869804664,
      "learning_rate": 8.979436164848088e-08,
      "loss": 1.0854,
      "step": 505
    },
    {
      "epoch": 0.9541627689429373,
      "grad_norm": 0.1538401535295636,
      "learning_rate": 6.15582970243117e-08,
      "loss": 1.0904,
      "step": 510
    },
    {
      "epoch": 0.9635173058933583,
      "grad_norm": 0.1590652953665931,
      "learning_rate": 3.861043947016474e-08,
      "loss": 1.0783,
      "step": 515
    },
    {
      "epoch": 0.9728718428437793,
      "grad_norm": 0.15600780093245614,
      "learning_rate": 2.097536212669171e-08,
      "loss": 1.0857,
      "step": 520
    },
    {
      "epoch": 0.9822263797942001,
      "grad_norm": 0.1590897713521175,
      "learning_rate": 8.671949076420883e-09,
      "loss": 1.0859,
      "step": 525
    },
    {
      "epoch": 0.9915809167446211,
      "grad_norm": 0.15861216238412335,
      "learning_rate": 1.7133751222137007e-09,
      "loss": 1.0845,
      "step": 530
    },
    {
      "epoch": 0.9990645463049579,
      "eval_runtime": 3.2564,
      "eval_samples_per_second": 3.071,
      "eval_steps_per_second": 0.921,
      "step": 534
    },
    {
      "epoch": 0.9990645463049579,
      "step": 534,
      "total_flos": 2.2276586722453094e+17,
      "train_loss": 1.1599709373734863,
      "train_runtime": 17757.3527,
      "train_samples_per_second": 1.925,
      "train_steps_per_second": 0.03
    }
  ],
  "logging_steps": 5,
  "max_steps": 534,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2.2276586722453094e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}