{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.895572263993316,
  "eval_steps": 500,
  "global_step": 2384,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3341687552213868,
      "grad_norm": 203.0,
      "learning_rate": 1.9596977329974813e-05,
      "loss": 3.5144,
      "step": 50
    },
    {
      "epoch": 0.6683375104427736,
      "grad_norm": 272.0,
      "learning_rate": 1.9177162048698573e-05,
      "loss": 3.2008,
      "step": 100
    },
    {
      "epoch": 1.0,
      "grad_norm": 205.0,
      "learning_rate": 1.8757346767422336e-05,
      "loss": 3.156,
      "step": 150
    },
    {
      "epoch": 1.3341687552213868,
      "grad_norm": 366.0,
      "learning_rate": 1.8337531486146097e-05,
      "loss": 2.6074,
      "step": 200
    },
    {
      "epoch": 1.6683375104427736,
      "grad_norm": 524.0,
      "learning_rate": 1.791771620486986e-05,
      "loss": 2.1476,
      "step": 250
    },
    {
      "epoch": 2.0,
      "grad_norm": 310.0,
      "learning_rate": 1.749790092359362e-05,
      "loss": 1.7122,
      "step": 300
    },
    {
      "epoch": 2.334168755221387,
      "grad_norm": 332.0,
      "learning_rate": 1.707808564231738e-05,
      "loss": 0.9868,
      "step": 350
    },
    {
      "epoch": 2.6683375104427736,
      "grad_norm": 264.0,
      "learning_rate": 1.6658270361041144e-05,
      "loss": 0.8583,
      "step": 400
    },
    {
      "epoch": 3.0,
      "grad_norm": 221.0,
      "learning_rate": 1.6238455079764907e-05,
      "loss": 0.6614,
      "step": 450
    },
    {
      "epoch": 3.334168755221387,
      "grad_norm": 247.0,
      "learning_rate": 1.5818639798488667e-05,
      "loss": 0.4849,
      "step": 500
    },
    {
      "epoch": 3.6683375104427736,
      "grad_norm": 156.0,
      "learning_rate": 1.5398824517212427e-05,
      "loss": 0.4501,
      "step": 550
    },
    {
      "epoch": 4.0,
      "grad_norm": 132.0,
      "learning_rate": 1.4979009235936189e-05,
      "loss": 0.4275,
      "step": 600
    },
    {
      "epoch": 4.334168755221387,
      "grad_norm": 218.0,
      "learning_rate": 1.455919395465995e-05,
      "loss": 0.3567,
      "step": 650
    },
    {
      "epoch": 4.668337510442774,
      "grad_norm": 129.0,
      "learning_rate": 1.4139378673383712e-05,
      "loss": 0.3326,
      "step": 700
    },
    {
      "epoch": 5.0,
      "grad_norm": 115.0,
      "learning_rate": 1.3719563392107474e-05,
      "loss": 0.3078,
      "step": 750
    },
    {
      "epoch": 5.334168755221387,
      "grad_norm": 121.0,
      "learning_rate": 1.3299748110831234e-05,
      "loss": 0.283,
      "step": 800
    },
    {
      "epoch": 5.668337510442774,
      "grad_norm": 156.0,
      "learning_rate": 1.2879932829554998e-05,
      "loss": 0.2888,
      "step": 850
    },
    {
      "epoch": 6.0,
      "grad_norm": 108.5,
      "learning_rate": 1.2460117548278758e-05,
      "loss": 0.2469,
      "step": 900
    },
    {
      "epoch": 6.334168755221387,
      "grad_norm": 151.0,
      "learning_rate": 1.204030226700252e-05,
      "loss": 0.2764,
      "step": 950
    },
    {
      "epoch": 6.668337510442774,
      "grad_norm": 105.5,
      "learning_rate": 1.162048698572628e-05,
      "loss": 0.2275,
      "step": 1000
    },
    {
      "epoch": 7.0,
      "grad_norm": 97.0,
      "learning_rate": 1.1200671704450043e-05,
      "loss": 0.1984,
      "step": 1050
    },
    {
      "epoch": 7.334168755221387,
      "grad_norm": 102.0,
      "learning_rate": 1.0780856423173805e-05,
      "loss": 0.218,
      "step": 1100
    },
    {
      "epoch": 7.668337510442774,
      "grad_norm": 108.0,
      "learning_rate": 1.0361041141897565e-05,
      "loss": 0.2004,
      "step": 1150
    },
    {
      "epoch": 8.0,
      "grad_norm": 64.5,
      "learning_rate": 9.941225860621327e-06,
      "loss": 0.1896,
      "step": 1200
    },
    {
      "epoch": 8.334168755221388,
      "grad_norm": 95.5,
      "learning_rate": 9.521410579345088e-06,
      "loss": 0.1903,
      "step": 1250
    },
    {
      "epoch": 8.668337510442774,
      "grad_norm": 101.0,
      "learning_rate": 9.10159529806885e-06,
      "loss": 0.1831,
      "step": 1300
    },
    {
      "epoch": 9.0,
      "grad_norm": 68.0,
      "learning_rate": 8.681780016792612e-06,
      "loss": 0.1624,
      "step": 1350
    },
    {
      "epoch": 9.334168755221388,
      "grad_norm": 83.5,
      "learning_rate": 8.261964735516374e-06,
      "loss": 0.1639,
      "step": 1400
    },
    {
      "epoch": 9.668337510442774,
      "grad_norm": 74.5,
      "learning_rate": 7.842149454240135e-06,
      "loss": 0.1834,
      "step": 1450
    },
    {
      "epoch": 10.0,
      "grad_norm": 35.5,
      "learning_rate": 7.422334172963896e-06,
      "loss": 0.1441,
      "step": 1500
    },
    {
      "epoch": 10.334168755221388,
      "grad_norm": 71.5,
      "learning_rate": 7.002518891687659e-06,
      "loss": 0.1547,
      "step": 1550
    },
    {
      "epoch": 10.668337510442774,
      "grad_norm": 85.5,
      "learning_rate": 6.58270361041142e-06,
      "loss": 0.1532,
      "step": 1600
    },
    {
      "epoch": 11.0,
      "grad_norm": 70.5,
      "learning_rate": 6.162888329135182e-06,
      "loss": 0.1452,
      "step": 1650
    },
    {
      "epoch": 11.334168755221388,
      "grad_norm": 83.5,
      "learning_rate": 5.7430730478589425e-06,
      "loss": 0.1294,
      "step": 1700
    },
    {
      "epoch": 11.668337510442774,
      "grad_norm": 49.75,
      "learning_rate": 5.323257766582704e-06,
      "loss": 0.167,
      "step": 1750
    },
    {
      "epoch": 12.0,
      "grad_norm": 52.25,
      "learning_rate": 4.903442485306465e-06,
      "loss": 0.1306,
      "step": 1800
    },
    {
      "epoch": 12.334168755221388,
      "grad_norm": 63.5,
      "learning_rate": 4.483627204030227e-06,
      "loss": 0.1376,
      "step": 1850
    },
    {
      "epoch": 12.668337510442774,
      "grad_norm": 71.5,
      "learning_rate": 4.063811922753989e-06,
      "loss": 0.1307,
      "step": 1900
    },
    {
      "epoch": 13.0,
      "grad_norm": 48.5,
      "learning_rate": 3.64399664147775e-06,
      "loss": 0.1369,
      "step": 1950
    },
    {
      "epoch": 13.334168755221388,
      "grad_norm": 91.0,
      "learning_rate": 3.2241813602015114e-06,
      "loss": 0.1284,
      "step": 2000
    },
    {
      "epoch": 13.668337510442774,
      "grad_norm": 82.5,
      "learning_rate": 2.804366078925273e-06,
      "loss": 0.1313,
      "step": 2050
    },
    {
      "epoch": 14.0,
      "grad_norm": 53.25,
      "learning_rate": 2.3845507976490345e-06,
      "loss": 0.1377,
      "step": 2100
    },
    {
      "epoch": 14.334168755221388,
      "grad_norm": 81.0,
      "learning_rate": 1.9647355163727962e-06,
      "loss": 0.1286,
      "step": 2150
    },
    {
      "epoch": 14.668337510442774,
      "grad_norm": 71.5,
      "learning_rate": 1.5449202350965576e-06,
      "loss": 0.1481,
      "step": 2200
    },
    {
      "epoch": 15.0,
      "grad_norm": 41.0,
      "learning_rate": 1.1251049538203191e-06,
      "loss": 0.1215,
      "step": 2250
    },
    {
      "epoch": 15.334168755221388,
      "grad_norm": 77.5,
      "learning_rate": 7.052896725440807e-07,
      "loss": 0.1247,
      "step": 2300
    },
    {
      "epoch": 15.668337510442774,
      "grad_norm": 70.0,
      "learning_rate": 2.8547439126784216e-07,
      "loss": 0.1307,
      "step": 2350
    }
  ],
  "logging_steps": 50,
  "max_steps": 2384,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 16,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.861877575078e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}