{
  "best_global_step": 293,
  "best_metric": 0.7508089542388916,
  "best_model_checkpoint": "/data/output/checkpoint-293",
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 293,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.9166833657771349,
      "epoch": 0.034158838599487616,
      "grad_norm": 3.222971200942993,
      "learning_rate": 0.00012,
      "loss": 0.9930056571960449,
      "mean_token_accuracy": 0.7560420431196689,
      "num_tokens": 527570.0,
      "step": 10
    },
    {
      "entropy": 0.8743617489933968,
      "epoch": 0.06831767719897523,
      "grad_norm": 2.950977325439453,
      "learning_rate": 0.00019989785286500295,
      "loss": 0.8596245765686035,
      "mean_token_accuracy": 0.7780317589640617,
      "num_tokens": 911738.0,
      "step": 20
    },
    {
      "entropy": 0.8505361303687096,
      "epoch": 0.10247651579846286,
      "grad_norm": 2.1713857650756836,
      "learning_rate": 0.00019875109281794825,
      "loss": 0.8519834518432617,
      "mean_token_accuracy": 0.7800005592405796,
      "num_tokens": 1217407.0,
      "step": 30
    },
    {
      "entropy": 0.8255588114261627,
      "epoch": 0.13663535439795046,
      "grad_norm": 2.044369697570801,
      "learning_rate": 0.00019634456691705702,
      "loss": 0.8329530715942383,
      "mean_token_accuracy": 0.7832169122993946,
      "num_tokens": 1459942.0,
      "step": 40
    },
    {
      "entropy": 0.7476669192314148,
      "epoch": 0.1707941929974381,
      "grad_norm": 2.332352638244629,
      "learning_rate": 0.00019270897516847403,
      "loss": 0.7421693325042724,
      "mean_token_accuracy": 0.8033282555639744,
      "num_tokens": 1607170.0,
      "step": 50
    },
    {
      "epoch": 0.1707941929974381,
      "eval_entropy": 0.7600614526867866,
      "eval_loss": 0.8239800930023193,
      "eval_mean_token_accuracy": 0.783122000694275,
      "eval_num_tokens": 1607170.0,
      "eval_runtime": 68.4494,
      "eval_samples_per_second": 1.461,
      "eval_steps_per_second": 1.461,
      "step": 50
    },
    {
      "entropy": 0.9212068691849709,
      "epoch": 0.2049530315969257,
      "grad_norm": 2.5008842945098877,
      "learning_rate": 0.0001878906967484966,
      "loss": 0.9536626815795899,
      "mean_token_accuracy": 0.7510255746543407,
      "num_tokens": 2140236.0,
      "step": 60
    },
    {
      "entropy": 0.892504171282053,
      "epoch": 0.23911187019641333,
      "grad_norm": 2.6001062393188477,
      "learning_rate": 0.00018195119834528534,
      "loss": 0.8938531875610352,
      "mean_token_accuracy": 0.7676444724202156,
      "num_tokens": 2529105.0,
      "step": 70
    },
    {
      "entropy": 0.8622478015720845,
      "epoch": 0.27327070879590093,
      "grad_norm": 2.4243602752685547,
      "learning_rate": 0.000174966250029467,
      "loss": 0.8663266181945801,
      "mean_token_accuracy": 0.7741704121232033,
      "num_tokens": 2827030.0,
      "step": 80
    },
    {
      "entropy": 0.8519950624555349,
      "epoch": 0.3074295473953886,
      "grad_norm": 2.669633150100708,
      "learning_rate": 0.0001670249586567531,
      "loss": 0.8582368850708008,
      "mean_token_accuracy": 0.7774012103676796,
      "num_tokens": 3056051.0,
      "step": 90
    },
    {
      "entropy": 0.7770494304597377,
      "epoch": 0.3415883859948762,
      "grad_norm": 2.272737741470337,
      "learning_rate": 0.00015822863113343935,
      "loss": 0.7583271026611328,
      "mean_token_accuracy": 0.7930335685610771,
      "num_tokens": 3197462.0,
      "step": 100
    },
    {
      "epoch": 0.3415883859948762,
      "eval_entropy": 0.7913119502365589,
      "eval_loss": 0.8178356885910034,
      "eval_mean_token_accuracy": 0.7806616699695588,
      "eval_num_tokens": 3197462.0,
      "eval_runtime": 68.4545,
      "eval_samples_per_second": 1.461,
      "eval_steps_per_second": 1.461,
      "step": 100
    },
    {
      "entropy": 0.8400638666003942,
      "epoch": 0.37574722459436377,
      "grad_norm": 2.6273744106292725,
      "learning_rate": 0.000148689482046087,
      "loss": 0.874265193939209,
      "mean_token_accuracy": 0.7818703927099705,
      "num_tokens": 3723005.0,
      "step": 110
    },
    {
      "entropy": 0.925657919421792,
      "epoch": 0.4099060631938514,
      "grad_norm": 2.1081862449645996,
      "learning_rate": 0.00013852920214212964,
      "loss": 0.9439982414245606,
      "mean_token_accuracy": 0.7651735462248326,
      "num_tokens": 4114734.0,
      "step": 120
    },
    {
      "entropy": 0.882326266542077,
      "epoch": 0.444064901793339,
      "grad_norm": 2.2644991874694824,
      "learning_rate": 0.0001278774059232723,
      "loss": 0.8633623123168945,
      "mean_token_accuracy": 0.7738826431334018,
      "num_tokens": 4422977.0,
      "step": 130
    },
    {
      "entropy": 0.807782543823123,
      "epoch": 0.47822374039282667,
      "grad_norm": 3.417691946029663,
      "learning_rate": 0.00011686997815570473,
      "loss": 0.8137018203735351,
      "mean_token_accuracy": 0.7846895553171634,
      "num_tokens": 4655796.0,
      "step": 140
    },
    {
      "entropy": 0.7050723437219858,
      "epoch": 0.5123825789923142,
      "grad_norm": 2.4027786254882812,
      "learning_rate": 0.00010564734039066699,
      "loss": 0.712303638458252,
      "mean_token_accuracy": 0.8001660093665123,
      "num_tokens": 4792918.0,
      "step": 150
    },
    {
      "epoch": 0.5123825789923142,
      "eval_entropy": 0.8103001257777214,
      "eval_loss": 0.7926951050758362,
      "eval_mean_token_accuracy": 0.7870555037260055,
      "eval_num_tokens": 4792918.0,
      "eval_runtime": 68.4439,
      "eval_samples_per_second": 1.461,
      "eval_steps_per_second": 1.461,
      "step": 150
    },
    {
      "entropy": 0.9130790904164314,
      "epoch": 0.5465414175918019,
      "grad_norm": 3.106920003890991,
      "learning_rate": 9.435265960933302e-05,
      "loss": 0.9007230758666992,
      "mean_token_accuracy": 0.765869815647602,
      "num_tokens": 5341793.0,
      "step": 160
    },
    {
      "entropy": 0.8538723237812519,
      "epoch": 0.5807002561912895,
      "grad_norm": 2.4355342388153076,
      "learning_rate": 8.313002184429529e-05,
      "loss": 0.8396717071533203,
      "mean_token_accuracy": 0.7828280851244926,
      "num_tokens": 5741381.0,
      "step": 170
    },
    {
      "entropy": 0.8008106090128422,
      "epoch": 0.6148590947907772,
      "grad_norm": 2.2028732299804688,
      "learning_rate": 7.21225940767277e-05,
      "loss": 0.8295732498168945,
      "mean_token_accuracy": 0.7857915386557579,
      "num_tokens": 6044669.0,
      "step": 180
    },
    {
      "entropy": 0.7763356593437493,
      "epoch": 0.6490179333902647,
      "grad_norm": 2.3838558197021484,
      "learning_rate": 6.147079785787038e-05,
      "loss": 0.7925190925598145,
      "mean_token_accuracy": 0.7925213143229485,
      "num_tokens": 6265758.0,
      "step": 190
    },
    {
      "entropy": 0.7516565401107073,
      "epoch": 0.6831767719897524,
      "grad_norm": 2.4081294536590576,
      "learning_rate": 5.131051795391302e-05,
      "loss": 0.7475024223327636,
      "mean_token_accuracy": 0.7955034308135509,
      "num_tokens": 6393481.0,
      "step": 200
    },
    {
      "epoch": 0.6831767719897524,
      "eval_entropy": 0.7843938270211219,
      "eval_loss": 0.7650399208068848,
      "eval_mean_token_accuracy": 0.7912229967117309,
      "eval_num_tokens": 6393481.0,
      "eval_runtime": 68.4585,
      "eval_samples_per_second": 1.461,
      "eval_steps_per_second": 1.461,
      "step": 200
    },
    {
      "entropy": 0.8729221720248461,
      "epoch": 0.71733561058924,
      "grad_norm": 2.095632791519165,
      "learning_rate": 4.1771368866560665e-05,
      "loss": 0.8615994453430176,
      "mean_token_accuracy": 0.7780913896858692,
      "num_tokens": 6941504.0,
      "step": 210
    },
    {
      "entropy": 0.8114416748285294,
      "epoch": 0.7514944491887275,
      "grad_norm": 2.0843183994293213,
      "learning_rate": 3.2975041343246936e-05,
      "loss": 0.8434048652648926,
      "mean_token_accuracy": 0.7824041344225406,
      "num_tokens": 7343592.0,
      "step": 220
    },
    {
      "entropy": 0.8621546909213066,
      "epoch": 0.7856532877882152,
      "grad_norm": 2.093738317489624,
      "learning_rate": 2.5033749970533015e-05,
      "loss": 0.8560552597045898,
      "mean_token_accuracy": 0.7798775337636471,
      "num_tokens": 7660705.0,
      "step": 230
    },
    {
      "entropy": 0.756019243132323,
      "epoch": 0.8198121263877028,
      "grad_norm": 2.397739887237549,
      "learning_rate": 1.8048801654714688e-05,
      "loss": 0.7541797161102295,
      "mean_token_accuracy": 0.801513460278511,
      "num_tokens": 7899243.0,
      "step": 240
    },
    {
      "entropy": 0.6868984286673367,
      "epoch": 0.8539709649871904,
      "grad_norm": 2.3398942947387695,
      "learning_rate": 1.2109303251503434e-05,
      "loss": 0.6575294971466065,
      "mean_token_accuracy": 0.8151952736079693,
      "num_tokens": 8040306.0,
      "step": 250
    },
    {
      "epoch": 0.8539709649871904,
      "eval_entropy": 0.761812056005001,
      "eval_loss": 0.7535683512687683,
      "eval_mean_token_accuracy": 0.7959340953826904,
      "eval_num_tokens": 8040306.0,
      "eval_runtime": 68.4452,
      "eval_samples_per_second": 1.461,
      "eval_steps_per_second": 1.461,
      "step": 250
    },
    {
      "entropy": 0.8774461716413497,
      "epoch": 0.888129803586678,
      "grad_norm": 3.395538091659546,
      "learning_rate": 7.291024831525961e-06,
      "loss": 0.8889847755432129,
      "mean_token_accuracy": 0.7704262785613537,
      "num_tokens": 524637.0,
      "step": 260
    },
    {
      "entropy": 0.8536802887916565,
      "epoch": 0.9222886421861657,
      "grad_norm": 2.4246485233306885,
      "learning_rate": 3.655433082942972e-06,
      "loss": 0.874198055267334,
      "mean_token_accuracy": 0.7646903492510319,
      "num_tokens": 896801.0,
      "step": 270
    },
    {
      "entropy": 0.8110974848270416,
      "epoch": 0.9564474807856533,
      "grad_norm": 2.450005531311035,
      "learning_rate": 1.2489071820517396e-06,
      "loss": 0.8041668891906738,
      "mean_token_accuracy": 0.7852038599550724,
      "num_tokens": 1182207.0,
      "step": 280
    },
    {
      "entropy": 0.7051055304706096,
      "epoch": 0.9906063193851409,
      "grad_norm": 2.2891805171966553,
      "learning_rate": 1.0214713499706597e-07,
      "loss": 0.710789155960083,
      "mean_token_accuracy": 0.80729488581419,
      "num_tokens": 1374112.0,
      "step": 290
    },
    {
      "epoch": 1.0,
      "eval_entropy": 0.7507871267944575,
      "eval_loss": 0.7508089542388916,
      "eval_mean_token_accuracy": 0.7960579419136047,
      "eval_num_tokens": 1401863.0,
      "eval_runtime": 68.5518,
      "eval_samples_per_second": 1.459,
      "eval_steps_per_second": 1.459,
      "step": 293
    }
  ],
  "logging_steps": 10,
  "max_steps": 293,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.0834807145130394e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}