{
  "best_global_step": 250,
  "best_metric": 0.9828886389732361,
  "best_model_checkpoint": "/data/output/checkpoint-250",
  "epoch": 0.8539709649871904,
  "eval_steps": 50,
  "global_step": 250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.1713112026453019,
      "epoch": 0.034158838599487616,
      "grad_norm": 3.3329014778137207,
      "learning_rate": 0.00012,
      "loss": 1.2999439239501953,
      "mean_token_accuracy": 0.6959523037075996,
      "num_tokens": 643010.0,
      "step": 10
    },
    {
      "entropy": 1.1543982975184917,
      "epoch": 0.06831767719897523,
      "grad_norm": 2.581118106842041,
      "learning_rate": 0.00019989785286500295,
      "loss": 1.1518776893615723,
      "mean_token_accuracy": 0.7155417971313,
      "num_tokens": 1251488.0,
      "step": 20
    },
    {
      "entropy": 0.9459363672882318,
      "epoch": 0.10247651579846286,
      "grad_norm": 2.3779942989349365,
      "learning_rate": 0.00019875109281794825,
      "loss": 0.9381013870239258,
      "mean_token_accuracy": 0.7682368114590645,
      "num_tokens": 1696074.0,
      "step": 30
    },
    {
      "entropy": 0.9385578982532025,
      "epoch": 0.13663535439795046,
      "grad_norm": 2.0882723331451416,
      "learning_rate": 0.00019634456691705702,
      "loss": 0.9547281265258789,
      "mean_token_accuracy": 0.7626850850880146,
      "num_tokens": 2006643.0,
      "step": 40
    },
    {
      "entropy": 0.9006230250000954,
      "epoch": 0.1707941929974381,
      "grad_norm": 2.151762008666992,
      "learning_rate": 0.00019270897516847403,
      "loss": 0.9329019546508789,
      "mean_token_accuracy": 0.7718432284891605,
      "num_tokens": 2177826.0,
      "step": 50
    },
    {
      "epoch": 0.1707941929974381,
      "eval_entropy": 1.0149640817940235,
      "eval_loss": 1.071570634841919,
      "eval_mean_token_accuracy": 0.742358745932579,
      "eval_num_tokens": 2177826.0,
      "eval_runtime": 95.085,
      "eval_samples_per_second": 1.052,
      "eval_steps_per_second": 1.052,
      "step": 50
    },
    {
      "entropy": 1.2094877127557992,
      "epoch": 0.2049530315969257,
      "grad_norm": 2.767357587814331,
      "learning_rate": 0.0001878906967484966,
      "loss": 1.209201717376709,
      "mean_token_accuracy": 0.7017950341105461,
      "num_tokens": 644405.0,
      "step": 60
    },
    {
      "entropy": 1.1481261406093837,
      "epoch": 0.23911187019641333,
      "grad_norm": 2.805593252182007,
      "learning_rate": 0.00018195119834528534,
      "loss": 1.1588325500488281,
      "mean_token_accuracy": 0.7221003979444504,
      "num_tokens": 1268539.0,
      "step": 70
    },
    {
      "entropy": 0.9842961743474007,
      "epoch": 0.27327070879590093,
      "grad_norm": 2.233400344848633,
      "learning_rate": 0.000174966250029467,
      "loss": 1.0100913047790527,
      "mean_token_accuracy": 0.7561626233160496,
      "num_tokens": 1729123.0,
      "step": 80
    },
    {
      "entropy": 0.9889072403311729,
      "epoch": 0.3074295473953886,
      "grad_norm": 1.992038369178772,
      "learning_rate": 0.0001670249586567531,
      "loss": 1.0143807411193848,
      "mean_token_accuracy": 0.7507779069244862,
      "num_tokens": 2034736.0,
      "step": 90
    },
    {
      "entropy": 0.8992462156340479,
      "epoch": 0.3415883859948762,
      "grad_norm": 2.0321593284606934,
      "learning_rate": 0.00015822863113343935,
      "loss": 0.9014181137084961,
      "mean_token_accuracy": 0.7751547530293464,
      "num_tokens": 2199862.0,
      "step": 100
    },
    {
      "epoch": 0.3415883859948762,
      "eval_entropy": 0.9943241326510907,
      "eval_loss": 1.043296456336975,
      "eval_mean_token_accuracy": 0.7519984298944473,
      "eval_num_tokens": 2199862.0,
      "eval_runtime": 95.5899,
      "eval_samples_per_second": 1.046,
      "eval_steps_per_second": 1.046,
      "step": 100
    },
    {
      "entropy": 1.1760564662516118,
      "epoch": 0.37574722459436377,
      "grad_norm": 3.0291037559509277,
      "learning_rate": 0.000148689482046087,
      "loss": 1.2000227928161622,
      "mean_token_accuracy": 0.711453115195036,
      "num_tokens": 643606.0,
      "step": 110
    },
    {
      "entropy": 1.1263522632420062,
      "epoch": 0.4099060631938514,
      "grad_norm": 2.554089307785034,
      "learning_rate": 0.00013852920214212964,
      "loss": 1.1538644790649415,
      "mean_token_accuracy": 0.7203516490757466,
      "num_tokens": 1250932.0,
      "step": 120
    },
    {
      "entropy": 0.9661639135330915,
      "epoch": 0.444064901793339,
      "grad_norm": 2.2706644535064697,
      "learning_rate": 0.0001278774059232723,
      "loss": 0.9473760604858399,
      "mean_token_accuracy": 0.7646104976534843,
      "num_tokens": 1690510.0,
      "step": 130
    },
    {
      "entropy": 0.9644285907968879,
      "epoch": 0.47822374039282667,
      "grad_norm": 1.8326361179351807,
      "learning_rate": 0.00011686997815570473,
      "loss": 0.9465847969055176,
      "mean_token_accuracy": 0.760813869535923,
      "num_tokens": 1991606.0,
      "step": 140
    },
    {
      "entropy": 0.847778935264796,
      "epoch": 0.5123825789923142,
      "grad_norm": 2.5428972244262695,
      "learning_rate": 0.00010564734039066699,
      "loss": 0.8805885314941406,
      "mean_token_accuracy": 0.784095474332571,
      "num_tokens": 2153462.0,
      "step": 150
    },
    {
      "epoch": 0.5123825789923142,
      "eval_entropy": 1.0146987434476613,
      "eval_loss": 1.0139193534851074,
      "eval_mean_token_accuracy": 0.7562771487236023,
      "eval_num_tokens": 2153462.0,
      "eval_runtime": 95.2112,
      "eval_samples_per_second": 1.05,
      "eval_steps_per_second": 1.05,
      "step": 150
    },
    {
      "entropy": 1.2566933140158654,
      "epoch": 0.5465414175918019,
      "grad_norm": 2.956472873687744,
      "learning_rate": 9.435265960933302e-05,
      "loss": 1.2643850326538086,
      "mean_token_accuracy": 0.700284918397665,
      "num_tokens": 644011.0,
      "step": 160
    },
    {
      "entropy": 1.1445974234491587,
      "epoch": 0.5807002561912895,
      "grad_norm": 2.522840738296509,
      "learning_rate": 8.313002184429529e-05,
      "loss": 1.1480578422546386,
      "mean_token_accuracy": 0.7231714740395546,
      "num_tokens": 1272288.0,
      "step": 170
    },
    {
      "entropy": 0.9980672411620617,
      "epoch": 0.6148590947907772,
      "grad_norm": 2.023250102996826,
      "learning_rate": 7.21225940767277e-05,
      "loss": 1.0411287307739259,
      "mean_token_accuracy": 0.74666218906641,
      "num_tokens": 1728820.0,
      "step": 180
    },
    {
      "entropy": 0.9762071590870619,
      "epoch": 0.6490179333902647,
      "grad_norm": 1.7645409107208252,
      "learning_rate": 6.147079785787038e-05,
      "loss": 0.9519403457641602,
      "mean_token_accuracy": 0.7596977911889553,
      "num_tokens": 2027911.0,
      "step": 190
    },
    {
      "entropy": 0.829610382579267,
      "epoch": 0.6831767719897524,
      "grad_norm": 1.8310364484786987,
      "learning_rate": 5.131051795391302e-05,
      "loss": 0.8207592964172363,
      "mean_token_accuracy": 0.7889963142573834,
      "num_tokens": 2200169.0,
      "step": 200
    },
    {
      "epoch": 0.6831767719897524,
      "eval_entropy": 1.0059119730442763,
      "eval_loss": 1.002184271812439,
      "eval_mean_token_accuracy": 0.7572826623916626,
      "eval_num_tokens": 2200169.0,
      "eval_runtime": 95.2716,
      "eval_samples_per_second": 1.05,
      "eval_steps_per_second": 1.05,
      "step": 200
    },
    {
      "entropy": 1.1494356069713831,
      "epoch": 0.71733561058924,
      "grad_norm": 3.3939480781555176,
      "learning_rate": 4.1771368866560665e-05,
      "loss": 1.169883918762207,
      "mean_token_accuracy": 0.7144785463809967,
      "num_tokens": 2843646.0,
      "step": 210
    },
    {
      "entropy": 1.1186032563447952,
      "epoch": 0.7514944491887275,
      "grad_norm": 2.491504192352295,
      "learning_rate": 3.2975041343246936e-05,
      "loss": 1.109489631652832,
      "mean_token_accuracy": 0.7285458266735076,
      "num_tokens": 3464241.0,
      "step": 220
    },
    {
      "entropy": 0.9699326818808913,
      "epoch": 0.7856532877882152,
      "grad_norm": 1.8113044500350952,
      "learning_rate": 2.5033749970533015e-05,
      "loss": 0.9755414009094239,
      "mean_token_accuracy": 0.7635111711919308,
      "num_tokens": 3914675.0,
      "step": 230
    },
    {
      "entropy": 0.9966869790107011,
      "epoch": 0.8198121263877028,
      "grad_norm": 1.7850583791732788,
      "learning_rate": 1.8048801654714688e-05,
      "loss": 1.020698356628418,
      "mean_token_accuracy": 0.7484274171292782,
      "num_tokens": 4212723.0,
      "step": 240
    },
    {
      "entropy": 0.8349324990995228,
      "epoch": 0.8539709649871904,
      "grad_norm": 1.9040881395339966,
      "learning_rate": 1.2109303251503434e-05,
      "loss": 0.8317201614379883,
      "mean_token_accuracy": 0.7891084231436253,
      "num_tokens": 4374805.0,
      "step": 250
    },
    {
      "epoch": 0.8539709649871904,
      "eval_entropy": 0.968962878882885,
      "eval_loss": 0.9828886389732361,
      "eval_mean_token_accuracy": 0.7627548807859421,
      "eval_num_tokens": 4374805.0,
      "eval_runtime": 95.2882,
      "eval_samples_per_second": 1.049,
      "eval_steps_per_second": 1.049,
      "step": 250
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.714933266799657e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}