{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.1749502982107356,
  "eval_steps": 500,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.9081946484744549,
      "epoch": 0.053015241882041084,
      "grad_norm": 51.5,
      "learning_rate": 3e-06,
      "loss": 2.320839500427246,
      "mean_token_accuracy": 0.5708254482597113,
      "num_tokens": 30761.0,
      "step": 10
    },
    {
      "entropy": 1.032745758444071,
      "epoch": 0.10603048376408217,
      "grad_norm": 21.125,
      "learning_rate": 6.333333333333333e-06,
      "loss": 1.9631355285644532,
      "mean_token_accuracy": 0.5809337079524994,
      "num_tokens": 61277.0,
      "step": 20
    },
    {
      "entropy": 1.2370933540165425,
      "epoch": 0.15904572564612326,
      "grad_norm": 11.0625,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.4653306961059571,
      "mean_token_accuracy": 0.6311864696443081,
      "num_tokens": 91730.0,
      "step": 30
    },
    {
      "entropy": 1.1646109029650689,
      "epoch": 0.21206096752816433,
      "grad_norm": 8.5625,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 1.200753402709961,
      "mean_token_accuracy": 0.6730783767998219,
      "num_tokens": 122896.0,
      "step": 40
    },
    {
      "entropy": 1.0138436160981654,
      "epoch": 0.26507620941020543,
      "grad_norm": 7.4375,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 1.0109354972839355,
      "mean_token_accuracy": 0.7080431424081326,
      "num_tokens": 153323.0,
      "step": 50
    },
    {
      "entropy": 0.8669569574296474,
      "epoch": 0.31809145129224653,
      "grad_norm": 6.75,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 0.8859419822692871,
      "mean_token_accuracy": 0.7356305047869682,
      "num_tokens": 183928.0,
      "step": 60
    },
    {
      "entropy": 0.856675973534584,
      "epoch": 0.3711066931742876,
      "grad_norm": 7.6875,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 0.8721119880676269,
      "mean_token_accuracy": 0.7318881675601006,
      "num_tokens": 215054.0,
      "step": 70
    },
    {
      "entropy": 0.7872001446783543,
      "epoch": 0.42412193505632867,
      "grad_norm": 7.46875,
      "learning_rate": 1.92962962962963e-05,
      "loss": 0.7920477390289307,
      "mean_token_accuracy": 0.7553018122911453,
      "num_tokens": 245614.0,
      "step": 80
    },
    {
      "entropy": 0.7739131577312947,
      "epoch": 0.47713717693836977,
      "grad_norm": 7.09375,
      "learning_rate": 1.8925925925925928e-05,
      "loss": 0.7949441432952881,
      "mean_token_accuracy": 0.7514781631529331,
      "num_tokens": 276651.0,
      "step": 90
    },
    {
      "entropy": 0.7763891488313674,
      "epoch": 0.5301524188204109,
      "grad_norm": 7.8125,
      "learning_rate": 1.8555555555555557e-05,
      "loss": 0.7867375373840332,
      "mean_token_accuracy": 0.7523738898336887,
      "num_tokens": 308561.0,
      "step": 100
    },
    {
      "entropy": 0.7757978036999702,
      "epoch": 0.583167660702452,
      "grad_norm": 7.21875,
      "learning_rate": 1.8185185185185186e-05,
      "loss": 0.7776416301727295,
      "mean_token_accuracy": 0.7548527352511882,
      "num_tokens": 339537.0,
      "step": 110
    },
    {
      "entropy": 0.7528716944158077,
      "epoch": 0.6361829025844931,
      "grad_norm": 6.28125,
      "learning_rate": 1.7814814814814815e-05,
      "loss": 0.7566296100616455,
      "mean_token_accuracy": 0.7637521453201771,
      "num_tokens": 370938.0,
      "step": 120
    },
    {
      "entropy": 0.7483146995306015,
      "epoch": 0.6891981444665342,
      "grad_norm": 7.0,
      "learning_rate": 1.7444444444444448e-05,
      "loss": 0.7508650302886963,
      "mean_token_accuracy": 0.758705198019743,
      "num_tokens": 402708.0,
      "step": 130
    },
    {
      "entropy": 0.7206243231892586,
      "epoch": 0.7422133863485753,
      "grad_norm": 6.65625,
      "learning_rate": 1.7074074074074077e-05,
      "loss": 0.7531259536743165,
      "mean_token_accuracy": 0.7641572073101998,
      "num_tokens": 433645.0,
      "step": 140
    },
    {
      "entropy": 0.7596748031675815,
      "epoch": 0.7952286282306164,
      "grad_norm": 6.03125,
      "learning_rate": 1.6703703703703703e-05,
      "loss": 0.7591583728790283,
      "mean_token_accuracy": 0.7570627965033054,
      "num_tokens": 465791.0,
      "step": 150
    },
    {
      "entropy": 0.7133973561227321,
      "epoch": 0.8482438701126573,
      "grad_norm": 7.125,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 0.7458220005035401,
      "mean_token_accuracy": 0.7626615799963474,
      "num_tokens": 497296.0,
      "step": 160
    },
    {
      "entropy": 0.7134327009320259,
      "epoch": 0.9012591119946984,
      "grad_norm": 7.0,
      "learning_rate": 1.5962962962962964e-05,
      "loss": 0.7290266990661621,
      "mean_token_accuracy": 0.7681502625346184,
      "num_tokens": 528181.0,
      "step": 170
    },
    {
      "entropy": 0.7324420973658562,
      "epoch": 0.9542743538767395,
      "grad_norm": 7.5,
      "learning_rate": 1.5592592592592593e-05,
      "loss": 0.731742525100708,
      "mean_token_accuracy": 0.7631207950413227,
      "num_tokens": 558915.0,
      "step": 180
    },
    {
      "entropy": 0.7100333733992144,
      "epoch": 1.0053015241882042,
      "grad_norm": 8.3125,
      "learning_rate": 1.5222222222222223e-05,
      "loss": 0.7303330898284912,
      "mean_token_accuracy": 0.7633489200047084,
      "num_tokens": 588567.0,
      "step": 190
    },
    {
      "entropy": 0.6628884050995112,
      "epoch": 1.0583167660702453,
      "grad_norm": 7.0625,
      "learning_rate": 1.4851851851851853e-05,
      "loss": 0.6481011867523193,
      "mean_token_accuracy": 0.7933171860873699,
      "num_tokens": 619081.0,
      "step": 200
    },
    {
      "entropy": 0.6257969986647367,
      "epoch": 1.1113320079522864,
      "grad_norm": 7.25,
      "learning_rate": 1.4481481481481483e-05,
      "loss": 0.642666244506836,
      "mean_token_accuracy": 0.7921255677938461,
      "num_tokens": 650402.0,
      "step": 210
    },
    {
      "entropy": 0.6399037420749665,
      "epoch": 1.1643472498343272,
      "grad_norm": 6.03125,
      "learning_rate": 1.4111111111111113e-05,
      "loss": 0.6525338172912598,
      "mean_token_accuracy": 0.7887163788080216,
      "num_tokens": 681837.0,
      "step": 220
    },
    {
      "entropy": 0.6367121540009976,
      "epoch": 1.2173624917163686,
      "grad_norm": 7.59375,
      "learning_rate": 1.3740740740740741e-05,
      "loss": 0.6356833457946778,
      "mean_token_accuracy": 0.7893597863614559,
      "num_tokens": 712571.0,
      "step": 230
    },
    {
      "entropy": 0.6334792256355286,
      "epoch": 1.2703777335984094,
      "grad_norm": 6.6875,
      "learning_rate": 1.3370370370370372e-05,
      "loss": 0.6391499996185303,
      "mean_token_accuracy": 0.7947010740637779,
      "num_tokens": 743396.0,
      "step": 240
    },
    {
      "entropy": 0.6417076423764229,
      "epoch": 1.3233929754804508,
      "grad_norm": 7.4375,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.6563210010528564,
      "mean_token_accuracy": 0.787147381901741,
      "num_tokens": 774883.0,
      "step": 250
    },
    {
      "entropy": 0.6213644035160542,
      "epoch": 1.3764082173624916,
      "grad_norm": 7.1875,
      "learning_rate": 1.2629629629629632e-05,
      "loss": 0.6295370578765869,
      "mean_token_accuracy": 0.7955153584480286,
      "num_tokens": 805520.0,
      "step": 260
    },
    {
      "entropy": 0.6552363462746144,
      "epoch": 1.4294234592445327,
      "grad_norm": 6.84375,
      "learning_rate": 1.225925925925926e-05,
      "loss": 0.6632986068725586,
      "mean_token_accuracy": 0.7826757229864597,
      "num_tokens": 836758.0,
      "step": 270
    },
    {
      "entropy": 0.6550090927630663,
      "epoch": 1.4824387011265738,
      "grad_norm": 7.25,
      "learning_rate": 1.188888888888889e-05,
      "loss": 0.6479949474334716,
      "mean_token_accuracy": 0.78950335085392,
      "num_tokens": 868056.0,
      "step": 280
    },
    {
      "entropy": 0.6275094959884882,
      "epoch": 1.535453943008615,
      "grad_norm": 6.4375,
      "learning_rate": 1.151851851851852e-05,
      "loss": 0.6384372234344482,
      "mean_token_accuracy": 0.7930210866034031,
      "num_tokens": 899420.0,
      "step": 290
    },
    {
      "entropy": 0.6052290312945843,
      "epoch": 1.588469184890656,
      "grad_norm": 6.875,
      "learning_rate": 1.114814814814815e-05,
      "loss": 0.609743309020996,
      "mean_token_accuracy": 0.7990770071744919,
      "num_tokens": 929571.0,
      "step": 300
    },
    {
      "entropy": 0.6174360021948815,
      "epoch": 1.6414844267726971,
      "grad_norm": 7.0625,
      "learning_rate": 1.0777777777777778e-05,
      "loss": 0.6313478946685791,
      "mean_token_accuracy": 0.7964761383831501,
      "num_tokens": 960928.0,
      "step": 310
    },
    {
      "entropy": 0.6127162992954254,
      "epoch": 1.6944996686547382,
      "grad_norm": 6.5625,
      "learning_rate": 1.0407407407407408e-05,
      "loss": 0.6150648593902588,
      "mean_token_accuracy": 0.8008938111364842,
      "num_tokens": 991961.0,
      "step": 320
    },
    {
      "entropy": 0.5947048265486956,
      "epoch": 1.7475149105367793,
      "grad_norm": 7.65625,
      "learning_rate": 1.003703703703704e-05,
      "loss": 0.6053301334381104,
      "mean_token_accuracy": 0.801536674797535,
      "num_tokens": 1022013.0,
      "step": 330
    },
    {
      "entropy": 0.6275908622890711,
      "epoch": 1.8005301524188204,
      "grad_norm": 6.9375,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.6307861804962158,
      "mean_token_accuracy": 0.7918523952364922,
      "num_tokens": 1053061.0,
      "step": 340
    },
    {
      "entropy": 0.6406017865985631,
      "epoch": 1.8535453943008615,
      "grad_norm": 7.0,
      "learning_rate": 9.296296296296296e-06,
      "loss": 0.6534273147583007,
      "mean_token_accuracy": 0.789715650677681,
      "num_tokens": 1085392.0,
      "step": 350
    },
    {
      "entropy": 0.6285702683031559,
      "epoch": 1.9065606361829026,
      "grad_norm": 7.8125,
      "learning_rate": 8.925925925925927e-06,
      "loss": 0.6331883907318115,
      "mean_token_accuracy": 0.7924636773765087,
      "num_tokens": 1117204.0,
      "step": 360
    },
    {
      "entropy": 0.615465309098363,
      "epoch": 1.9595758780649437,
      "grad_norm": 7.25,
      "learning_rate": 8.555555555555556e-06,
      "loss": 0.6278977870941163,
      "mean_token_accuracy": 0.7965366452932358,
      "num_tokens": 1147711.0,
      "step": 370
    },
    {
      "entropy": 0.6297557744886968,
      "epoch": 2.0106030483764084,
      "grad_norm": 6.46875,
      "learning_rate": 8.185185185185187e-06,
      "loss": 0.6424172878265381,
      "mean_token_accuracy": 0.7955868112576472,
      "num_tokens": 1177908.0,
      "step": 380
    },
    {
      "entropy": 0.599795724824071,
      "epoch": 2.0636182902584492,
      "grad_norm": 6.4375,
      "learning_rate": 7.814814814814816e-06,
      "loss": 0.5880236625671387,
      "mean_token_accuracy": 0.805521210283041,
      "num_tokens": 1209330.0,
      "step": 390
    },
    {
      "entropy": 0.5943233758211136,
      "epoch": 2.1166335321404905,
      "grad_norm": 6.9375,
      "learning_rate": 7.444444444444445e-06,
      "loss": 0.5809715270996094,
      "mean_token_accuracy": 0.8090382158756256,
      "num_tokens": 1241043.0,
      "step": 400
    },
    {
      "entropy": 0.583736015111208,
      "epoch": 2.1696487740225314,
      "grad_norm": 6.875,
      "learning_rate": 7.074074074074074e-06,
      "loss": 0.595209789276123,
      "mean_token_accuracy": 0.8054515205323696,
      "num_tokens": 1272875.0,
      "step": 410
    },
    {
      "entropy": 0.5891257628798485,
      "epoch": 2.2226640159045727,
      "grad_norm": 7.875,
      "learning_rate": 6.703703703703704e-06,
      "loss": 0.6067119598388672,
      "mean_token_accuracy": 0.8018631525337696,
      "num_tokens": 1303473.0,
      "step": 420
    },
    {
      "entropy": 0.5886123344302178,
      "epoch": 2.2756792577866136,
      "grad_norm": 7.65625,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.577965784072876,
      "mean_token_accuracy": 0.8081556148827076,
      "num_tokens": 1335327.0,
      "step": 430
    },
    {
      "entropy": 0.5796562805771828,
      "epoch": 2.3286944996686545,
      "grad_norm": 8.125,
      "learning_rate": 5.962962962962963e-06,
      "loss": 0.5881845951080322,
      "mean_token_accuracy": 0.8047773309051991,
      "num_tokens": 1366019.0,
      "step": 440
    },
    {
      "entropy": 0.590113490819931,
      "epoch": 2.381709741550696,
      "grad_norm": 6.96875,
      "learning_rate": 5.5925925925925926e-06,
      "loss": 0.593454647064209,
      "mean_token_accuracy": 0.809637114405632,
      "num_tokens": 1397101.0,
      "step": 450
    },
    {
      "entropy": 0.570321362093091,
      "epoch": 2.434724983432737,
      "grad_norm": 7.34375,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.5790628910064697,
      "mean_token_accuracy": 0.809483564645052,
      "num_tokens": 1428115.0,
      "step": 460
    },
    {
      "entropy": 0.6001236628741026,
      "epoch": 2.487740225314778,
      "grad_norm": 7.21875,
      "learning_rate": 4.851851851851852e-06,
      "loss": 0.6089425086975098,
      "mean_token_accuracy": 0.8031827084720135,
      "num_tokens": 1460337.0,
      "step": 470
    },
    {
      "entropy": 0.5818770505487919,
      "epoch": 2.540755467196819,
      "grad_norm": 6.5625,
      "learning_rate": 4.481481481481482e-06,
      "loss": 0.5913994789123536,
      "mean_token_accuracy": 0.8058472029864788,
      "num_tokens": 1491465.0,
      "step": 480
    },
    {
      "entropy": 0.5761014901101589,
      "epoch": 2.59377070907886,
      "grad_norm": 7.25,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.573754072189331,
      "mean_token_accuracy": 0.8119833193719387,
      "num_tokens": 1522283.0,
      "step": 490
    },
    {
      "entropy": 0.5881835252046586,
      "epoch": 2.6467859509609015,
      "grad_norm": 7.8125,
      "learning_rate": 3.740740740740741e-06,
      "loss": 0.5855993270874024,
      "mean_token_accuracy": 0.8061533674597741,
      "num_tokens": 1553066.0,
      "step": 500
    },
    {
      "entropy": 0.5923898722976446,
      "epoch": 2.6998011928429424,
      "grad_norm": 7.25,
      "learning_rate": 3.3703703703703705e-06,
      "loss": 0.5783270835876465,
      "mean_token_accuracy": 0.8092506147921086,
      "num_tokens": 1583785.0,
      "step": 510
    },
    {
      "entropy": 0.5890947930514813,
      "epoch": 2.7528164347249833,
      "grad_norm": 7.09375,
      "learning_rate": 3e-06,
      "loss": 0.5847487449645996,
      "mean_token_accuracy": 0.8107674546539784,
      "num_tokens": 1615166.0,
      "step": 520
    },
    {
      "entropy": 0.5714535418897867,
      "epoch": 2.8058316766070246,
      "grad_norm": 7.0,
      "learning_rate": 2.6296296296296297e-06,
      "loss": 0.5737119197845459,
      "mean_token_accuracy": 0.8139248043298721,
      "num_tokens": 1646166.0,
      "step": 530
    },
    {
      "entropy": 0.5585082449018955,
      "epoch": 2.8588469184890655,
      "grad_norm": 8.75,
      "learning_rate": 2.2592592592592592e-06,
      "loss": 0.5831988334655762,
      "mean_token_accuracy": 0.8078319534659386,
      "num_tokens": 1675970.0,
      "step": 540
    },
    {
      "entropy": 0.5816058319061994,
      "epoch": 2.911862160371107,
      "grad_norm": 6.6875,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.585014820098877,
      "mean_token_accuracy": 0.8070621207356453,
      "num_tokens": 1707365.0,
      "step": 550
    },
    {
      "entropy": 0.5718538902699948,
      "epoch": 2.9648774022531477,
      "grad_norm": 7.4375,
      "learning_rate": 1.5185185185185186e-06,
      "loss": 0.5958520889282226,
      "mean_token_accuracy": 0.8062198340892792,
      "num_tokens": 1737306.0,
      "step": 560
    },
    {
      "entropy": 0.5680852636888429,
      "epoch": 3.0159045725646125,
      "grad_norm": 7.28125,
      "learning_rate": 1.1481481481481482e-06,
      "loss": 0.5883944511413575,
      "mean_token_accuracy": 0.8095385347093854,
      "num_tokens": 1766093.0,
      "step": 570
    },
    {
      "entropy": 0.5735234707593918,
      "epoch": 3.0689198144466534,
      "grad_norm": 6.1875,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.5706701755523682,
      "mean_token_accuracy": 0.8123805336654186,
      "num_tokens": 1796844.0,
      "step": 580
    },
    {
      "entropy": 0.575826159119606,
      "epoch": 3.1219350563286943,
      "grad_norm": 6.9375,
      "learning_rate": 4.074074074074075e-07,
      "loss": 0.5520034790039062,
      "mean_token_accuracy": 0.8180661156773568,
      "num_tokens": 1828137.0,
      "step": 590
    },
    {
      "entropy": 0.5985390111804009,
      "epoch": 3.1749502982107356,
      "grad_norm": 7.34375,
      "learning_rate": 3.703703703703704e-08,
      "loss": 0.5971494674682617,
      "mean_token_accuracy": 0.8037057816982269,
      "num_tokens": 1859725.0,
      "step": 600
    }
  ],
  "logging_steps": 10,
  "max_steps": 600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 1.7956887511474176e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}