{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 874,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02288329519450801,
      "grad_norm": 18.221216778257453,
      "learning_rate": 1.0227272727272729e-06,
      "loss": 1.393819236755371,
      "step": 10
    },
    {
      "epoch": 0.04576659038901602,
      "grad_norm": 14.224119847108293,
      "learning_rate": 2.1590909090909092e-06,
      "loss": 1.384445571899414,
      "step": 20
    },
    {
      "epoch": 0.06864988558352403,
      "grad_norm": 8.084131033987092,
      "learning_rate": 3.2954545454545456e-06,
      "loss": 1.3395289421081542,
      "step": 30
    },
    {
      "epoch": 0.09153318077803203,
      "grad_norm": 4.825162041485144,
      "learning_rate": 4.4318181818181824e-06,
      "loss": 1.199374008178711,
      "step": 40
    },
    {
      "epoch": 0.11441647597254005,
      "grad_norm": 2.620746824810217,
      "learning_rate": 5.568181818181818e-06,
      "loss": 1.1844324111938476,
      "step": 50
    },
    {
      "epoch": 0.13729977116704806,
      "grad_norm": 2.280561836514811,
      "learning_rate": 6.704545454545454e-06,
      "loss": 1.1506938934326172,
      "step": 60
    },
    {
      "epoch": 0.16018306636155608,
      "grad_norm": 1.840017749876338,
      "learning_rate": 7.840909090909091e-06,
      "loss": 1.1342477798461914,
      "step": 70
    },
    {
      "epoch": 0.18306636155606407,
      "grad_norm": 1.4046794648896719,
      "learning_rate": 8.977272727272727e-06,
      "loss": 1.0628189086914062,
      "step": 80
    },
    {
      "epoch": 0.20594965675057209,
      "grad_norm": 1.1261167433814985,
      "learning_rate": 9.999960061285353e-06,
      "loss": 1.055792236328125,
      "step": 90
    },
    {
      "epoch": 0.2288329519450801,
      "grad_norm": 1.1859981099000974,
      "learning_rate": 9.995168187508393e-06,
      "loss": 1.0445270538330078,
      "step": 100
    },
    {
      "epoch": 0.2517162471395881,
      "grad_norm": 1.050590773848972,
      "learning_rate": 9.98239734155262e-06,
      "loss": 1.0704252243041992,
      "step": 110
    },
    {
      "epoch": 0.2745995423340961,
      "grad_norm": 0.9556182185508427,
      "learning_rate": 9.961667922776125e-06,
      "loss": 1.0515583992004394,
      "step": 120
    },
    {
      "epoch": 0.2974828375286041,
      "grad_norm": 1.0596424685797112,
      "learning_rate": 9.93301304306814e-06,
      "loss": 1.05296049118042,
      "step": 130
    },
    {
      "epoch": 0.32036613272311215,
      "grad_norm": 0.9934624870737894,
      "learning_rate": 9.896478473958147e-06,
      "loss": 1.0407039642333984,
      "step": 140
    },
    {
      "epoch": 0.34324942791762014,
      "grad_norm": 0.9296921808311289,
      "learning_rate": 9.852122573503283e-06,
      "loss": 1.015059471130371,
      "step": 150
    },
    {
      "epoch": 0.36613272311212813,
      "grad_norm": 1.0926612464245933,
      "learning_rate": 9.800016193070789e-06,
      "loss": 1.0180091857910156,
      "step": 160
    },
    {
      "epoch": 0.3890160183066362,
      "grad_norm": 0.9560875189280538,
      "learning_rate": 9.740242564164433e-06,
      "loss": 1.060804843902588,
      "step": 170
    },
    {
      "epoch": 0.41189931350114417,
      "grad_norm": 1.0082190996618579,
      "learning_rate": 9.672897165475657e-06,
      "loss": 0.9985034942626954,
      "step": 180
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.9293505463478718,
      "learning_rate": 9.59808757037183e-06,
      "loss": 1.0597731590270996,
      "step": 190
    },
    {
      "epoch": 0.4576659038901602,
      "grad_norm": 0.9174121034502059,
      "learning_rate": 9.515933275065218e-06,
      "loss": 1.024726390838623,
      "step": 200
    },
    {
      "epoch": 0.4805491990846682,
      "grad_norm": 0.8785467635869065,
      "learning_rate": 9.426565507737139e-06,
      "loss": 1.046273422241211,
      "step": 210
    },
    {
      "epoch": 0.5034324942791762,
      "grad_norm": 0.9212480009396277,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.9727144241333008,
      "step": 220
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.9164177333802558,
      "learning_rate": 9.226771853487411e-06,
      "loss": 1.0013788223266602,
      "step": 230
    },
    {
      "epoch": 0.5491990846681922,
      "grad_norm": 1.0380197289827424,
      "learning_rate": 9.116665104570513e-06,
      "loss": 1.0090283393859862,
      "step": 240
    },
    {
      "epoch": 0.5720823798627003,
      "grad_norm": 1.0045067275254234,
      "learning_rate": 8.999982649870377e-06,
      "loss": 1.0333093643188476,
      "step": 250
    },
    {
      "epoch": 0.5949656750572082,
      "grad_norm": 1.0842401366028722,
      "learning_rate": 8.876910870710885e-06,
      "loss": 1.0343000411987304,
      "step": 260
    },
    {
      "epoch": 0.6178489702517163,
      "grad_norm": 1.0444944619449643,
      "learning_rate": 8.747646354326948e-06,
      "loss": 1.0472993850708008,
      "step": 270
    },
    {
      "epoch": 0.6407322654462243,
      "grad_norm": 0.8966161743607417,
      "learning_rate": 8.612395579848244e-06,
      "loss": 1.007988166809082,
      "step": 280
    },
    {
      "epoch": 0.6636155606407322,
      "grad_norm": 1.0512223114619048,
      "learning_rate": 8.47137458848224e-06,
      "loss": 1.014136505126953,
      "step": 290
    },
    {
      "epoch": 0.6864988558352403,
      "grad_norm": 0.909017241303456,
      "learning_rate": 8.324808638423353e-06,
      "loss": 1.047772216796875,
      "step": 300
    },
    {
      "epoch": 0.7093821510297483,
      "grad_norm": 0.891327607364882,
      "learning_rate": 8.172931845039471e-06,
      "loss": 0.9750609397888184,
      "step": 310
    },
    {
      "epoch": 0.7322654462242563,
      "grad_norm": 0.9733547314088673,
      "learning_rate": 8.01598680691057e-06,
      "loss": 0.9990760803222656,
      "step": 320
    },
    {
      "epoch": 0.7551487414187643,
      "grad_norm": 0.9568324370937283,
      "learning_rate": 7.854224218316761e-06,
      "loss": 1.0234622955322266,
      "step": 330
    },
    {
      "epoch": 0.7780320366132724,
      "grad_norm": 1.2132080107449337,
      "learning_rate": 7.687902468794803e-06,
      "loss": 1.031675434112549,
      "step": 340
    },
    {
      "epoch": 0.8009153318077803,
      "grad_norm": 0.8978424025587594,
      "learning_rate": 7.517287230402639e-06,
      "loss": 0.9884134292602539,
      "step": 350
    },
    {
      "epoch": 0.8237986270022883,
      "grad_norm": 1.0233402755425876,
      "learning_rate": 7.342651033351325e-06,
      "loss": 0.9715648651123047,
      "step": 360
    },
    {
      "epoch": 0.8466819221967964,
      "grad_norm": 0.8907463390137706,
      "learning_rate": 7.164272830682161e-06,
      "loss": 1.0005844116210938,
      "step": 370
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.9992015638804682,
      "learning_rate": 6.9824375526843705e-06,
      "loss": 1.0225093841552735,
      "step": 380
    },
    {
      "epoch": 0.8924485125858124,
      "grad_norm": 0.9423403157154376,
      "learning_rate": 6.797435651765123e-06,
      "loss": 1.0305269241333008,
      "step": 390
    },
    {
      "epoch": 0.9153318077803204,
      "grad_norm": 0.9312149542771266,
      "learning_rate": 6.609562638498845e-06,
      "loss": 0.965084171295166,
      "step": 400
    },
    {
      "epoch": 0.9382151029748284,
      "grad_norm": 0.8994648058485575,
      "learning_rate": 6.419118609596948e-06,
      "loss": 0.9787817001342773,
      "step": 410
    },
    {
      "epoch": 0.9610983981693364,
      "grad_norm": 0.9292959943014952,
      "learning_rate": 6.2264077685519246e-06,
      "loss": 1.0170610427856446,
      "step": 420
    },
    {
      "epoch": 0.9839816933638444,
      "grad_norm": 1.0057392492482709,
      "learning_rate": 6.031737939721538e-06,
      "loss": 1.024329948425293,
      "step": 430
    },
    {
      "epoch": 1.0068649885583525,
      "grad_norm": 1.0121140233740662,
      "learning_rate": 5.835420076629273e-06,
      "loss": 1.0376602172851563,
      "step": 440
    },
    {
      "epoch": 1.0297482837528604,
      "grad_norm": 0.8930268980879296,
      "learning_rate": 5.6377677652664345e-06,
      "loss": 0.982276439666748,
      "step": 450
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.9716965280324289,
      "learning_rate": 5.439096723189316e-06,
      "loss": 1.0728231430053712,
      "step": 460
    },
    {
      "epoch": 1.0755148741418765,
      "grad_norm": 0.9065141249929061,
      "learning_rate": 5.239724295211541e-06,
      "loss": 1.0191450119018555,
      "step": 470
    },
    {
      "epoch": 1.0983981693363845,
      "grad_norm": 0.8012852118037016,
      "learning_rate": 5.03996894649711e-06,
      "loss": 1.0118200302124023,
      "step": 480
    },
    {
      "epoch": 1.1212814645308924,
      "grad_norm": 0.9890053077086055,
      "learning_rate": 4.840149753863887e-06,
      "loss": 0.9734249114990234,
      "step": 490
    },
    {
      "epoch": 1.1441647597254005,
      "grad_norm": 0.9130820112985458,
      "learning_rate": 4.640585896110054e-06,
      "loss": 1.0037405014038085,
      "step": 500
    },
    {
      "epoch": 1.1670480549199085,
      "grad_norm": 0.8720442088124731,
      "learning_rate": 4.441596144177681e-06,
      "loss": 1.0057022094726562,
      "step": 510
    },
    {
      "epoch": 1.1899313501144164,
      "grad_norm": 0.8993378778714946,
      "learning_rate": 4.243498351967761e-06,
      "loss": 0.9841510772705078,
      "step": 520
    },
    {
      "epoch": 1.2128146453089246,
      "grad_norm": 1.009853776998259,
      "learning_rate": 4.046608948620098e-06,
      "loss": 0.9782192230224609,
      "step": 530
    },
    {
      "epoch": 1.2356979405034325,
      "grad_norm": 0.9221177344151149,
      "learning_rate": 3.8512424330689855e-06,
      "loss": 0.9990880966186524,
      "step": 540
    },
    {
      "epoch": 1.2585812356979404,
      "grad_norm": 1.0454378267485458,
      "learning_rate": 3.657710871682115e-06,
      "loss": 1.003445529937744,
      "step": 550
    },
    {
      "epoch": 1.2814645308924484,
      "grad_norm": 0.8805351505795899,
      "learning_rate": 3.466323399785072e-06,
      "loss": 0.9950039863586426,
      "step": 560
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 0.9440338584576333,
      "learning_rate": 3.2773857278677414e-06,
      "loss": 1.0134332656860352,
      "step": 570
    },
    {
      "epoch": 1.3272311212814645,
      "grad_norm": 0.9837635767810521,
      "learning_rate": 3.0911996532612966e-06,
      "loss": 1.0345746994018554,
      "step": 580
    },
    {
      "epoch": 1.3501144164759724,
      "grad_norm": 1.0390562356673976,
      "learning_rate": 2.9080625780658455e-06,
      "loss": 0.9831764221191406,
      "step": 590
    },
    {
      "epoch": 1.3729977116704806,
      "grad_norm": 0.8741142160541207,
      "learning_rate": 2.7282670340987465e-06,
      "loss": 0.971280860900879,
      "step": 600
    },
    {
      "epoch": 1.3958810068649885,
      "grad_norm": 1.0166482283158509,
      "learning_rate": 2.552100215622401e-06,
      "loss": 1.0130719184875487,
      "step": 610
    },
    {
      "epoch": 1.4187643020594964,
      "grad_norm": 1.0532152530341887,
      "learning_rate": 2.379843520597937e-06,
      "loss": 1.0302873611450196,
      "step": 620
    },
    {
      "epoch": 1.4416475972540046,
      "grad_norm": 0.8906594744109658,
      "learning_rate": 2.2117721011975366e-06,
      "loss": 1.044809055328369,
      "step": 630
    },
    {
      "epoch": 1.4645308924485125,
      "grad_norm": 0.9182197471145602,
      "learning_rate": 2.0481544242934105e-06,
      "loss": 1.0100820541381836,
      "step": 640
    },
    {
      "epoch": 1.4874141876430205,
      "grad_norm": 0.9591797893121556,
      "learning_rate": 1.8892518426254363e-06,
      "loss": 0.9954021453857422,
      "step": 650
    },
    {
      "epoch": 1.5102974828375286,
      "grad_norm": 0.9536465128393342,
      "learning_rate": 1.735318177332503e-06,
      "loss": 1.0388307571411133,
      "step": 660
    },
    {
      "epoch": 1.5331807780320366,
      "grad_norm": 0.9851344067069513,
      "learning_rate": 1.5865993125143297e-06,
      "loss": 1.023411178588867,
      "step": 670
    },
    {
      "epoch": 1.5560640732265445,
      "grad_norm": 0.9120096357238873,
      "learning_rate": 1.4433328024714583e-06,
      "loss": 1.0003677368164063,
      "step": 680
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.8397516139080035,
      "learning_rate": 1.3057474922507124e-06,
      "loss": 0.9492273330688477,
      "step": 690
    },
    {
      "epoch": 1.6018306636155606,
      "grad_norm": 0.9803332570244476,
      "learning_rate": 1.1740631521023176e-06,
      "loss": 1.0218355178833007,
      "step": 700
    },
    {
      "epoch": 1.6247139588100685,
      "grad_norm": 1.0071220717489042,
      "learning_rate": 1.0484901264325026e-06,
      "loss": 1.0344181060791016,
      "step": 710
    },
    {
      "epoch": 1.6475972540045767,
      "grad_norm": 0.8809638651282645,
      "learning_rate": 9.292289978123892e-07,
      "loss": 0.945091438293457,
      "step": 720
    },
    {
      "epoch": 1.6704805491990846,
      "grad_norm": 1.0138393788661297,
      "learning_rate": 8.16470266579813e-07,
      "loss": 1.0206674575805663,
      "step": 730
    },
    {
      "epoch": 1.6933638443935926,
      "grad_norm": 0.9183367430377967,
      "learning_rate": 7.103940465458936e-07,
      "loss": 1.003372573852539,
      "step": 740
    },
    {
      "epoch": 1.7162471395881007,
      "grad_norm": 0.9808017585038956,
      "learning_rate": 6.111697772923836e-07,
      "loss": 1.002475643157959,
      "step": 750
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.8875163328550894,
      "learning_rate": 5.189559535193839e-07,
      "loss": 0.9872116088867188,
      "step": 760
    },
    {
      "epoch": 1.7620137299771166,
      "grad_norm": 1.0241092820794546,
      "learning_rate": 4.338998718757315e-07,
      "loss": 0.9817548751831054,
      "step": 770
    },
    {
      "epoch": 1.7848970251716247,
      "grad_norm": 0.9245308929706272,
      "learning_rate": 3.5613739567645246e-07,
      "loss": 1.0520426750183105,
      "step": 780
    },
    {
      "epoch": 1.8077803203661327,
      "grad_norm": 0.9500733954150005,
      "learning_rate": 2.8579273788312433e-07,
      "loss": 1.01949462890625,
      "step": 790
    },
    {
      "epoch": 1.8306636155606406,
      "grad_norm": 0.9550573822974243,
      "learning_rate": 2.2297826269378653e-07,
      "loss": 0.9677618026733399,
      "step": 800
    },
    {
      "epoch": 1.8535469107551488,
      "grad_norm": 0.8418605949321986,
      "learning_rate": 1.6779430605933233e-07,
      "loss": 1.0336989402770995,
      "step": 810
    },
    {
      "epoch": 1.8764302059496567,
      "grad_norm": 0.9224184199250552,
      "learning_rate": 1.203290154130754e-07,
      "loss": 1.053194236755371,
      "step": 820
    },
    {
      "epoch": 1.8993135011441646,
      "grad_norm": 0.8702703264528128,
      "learning_rate": 8.065820886950404e-08,
      "loss": 0.9445564270019531,
      "step": 830
    },
    {
      "epoch": 1.9221967963386728,
      "grad_norm": 0.8676834240948701,
      "learning_rate": 4.8845254117120265e-08,
      "loss": 1.0015798568725587,
      "step": 840
    },
    {
      "epoch": 1.9450800915331807,
      "grad_norm": 1.0429643813275697,
      "learning_rate": 2.4940967198821553e-08,
      "loss": 0.9959196090698242,
      "step": 850
    },
    {
      "epoch": 1.9679633867276887,
      "grad_norm": 0.9433672959674375,
      "learning_rate": 8.983531341500984e-09,
      "loss": 0.9756921768188477,
      "step": 860
    },
    {
      "epoch": 1.9908466819221968,
      "grad_norm": 1.0526736161532493,
      "learning_rate": 9.984359645276443e-10,
      "loss": 0.9688234329223633,
      "step": 870
    },
    {
      "epoch": 2.0,
      "step": 874,
      "total_flos": 2.312523470065369e+18,
      "train_loss": 1.0307597314738572,
      "train_runtime": 2075.8497,
      "train_samples_per_second": 53.887,
      "train_steps_per_second": 0.421
    }
  ],
  "logging_steps": 10,
  "max_steps": 874,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.312523470065369e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}