{
  "best_metric": 0.09521909803152084,
  "best_model_checkpoint": "outputs/checkpoint-329",
  "epoch": 5.995444191343964,
  "eval_steps": 500,
  "global_step": 329,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36446469248291574,
      "grad_norm": 3.342350721359253,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.3727,
      "step": 20
    },
    {
      "epoch": 0.7289293849658315,
      "grad_norm": 1.456776738166809,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.2857,
      "step": 40
    },
    {
      "epoch": 0.9840546697038725,
      "eval_loss": 0.1871582418680191,
      "eval_runtime": 48.9194,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 54
    },
    {
      "epoch": 1.0933940774487472,
      "grad_norm": 0.6021007895469666,
      "learning_rate": 5.8e-06,
      "loss": 0.1809,
      "step": 60
    },
    {
      "epoch": 1.4578587699316627,
      "grad_norm": 0.4066278040409088,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.1344,
      "step": 80
    },
    {
      "epoch": 1.8223234624145785,
      "grad_norm": 1.681560754776001,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.1242,
      "step": 100
    },
    {
      "epoch": 1.9863325740318907,
      "eval_loss": 0.1359640508890152,
      "eval_runtime": 48.8699,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.43,
      "step": 109
    },
    {
      "epoch": 2.1867881548974943,
      "grad_norm": 0.459069162607193,
      "learning_rate": 9.984149663879994e-06,
      "loss": 0.1039,
      "step": 120
    },
    {
      "epoch": 2.55125284738041,
      "grad_norm": 0.594571590423584,
      "learning_rate": 9.929487384240103e-06,
      "loss": 0.0874,
      "step": 140
    },
    {
      "epoch": 2.9157175398633255,
      "grad_norm": 0.40587714314460754,
      "learning_rate": 9.83624518217252e-06,
      "loss": 0.0862,
      "step": 160
    },
    {
      "epoch": 2.988610478359909,
      "eval_loss": 0.11714969575405121,
      "eval_runtime": 48.9241,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 164
    },
    {
      "epoch": 3.2801822323462413,
      "grad_norm": 0.5911823511123657,
      "learning_rate": 9.705152804330872e-06,
      "loss": 0.0833,
      "step": 180
    },
    {
      "epoch": 3.644646924829157,
      "grad_norm": 0.6380645036697388,
      "learning_rate": 9.53723622631339e-06,
      "loss": 0.0754,
      "step": 200
    },
    {
      "epoch": 3.990888382687927,
      "eval_loss": 0.10540632158517838,
      "eval_runtime": 48.8751,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 219
    },
    {
      "epoch": 4.009111617312073,
      "grad_norm": 0.5483060479164124,
      "learning_rate": 9.333809623012763e-06,
      "loss": 0.0731,
      "step": 220
    },
    {
      "epoch": 4.373576309794989,
      "grad_norm": 0.5601429343223572,
      "learning_rate": 9.096465083415809e-06,
      "loss": 0.0679,
      "step": 240
    },
    {
      "epoch": 4.738041002277904,
      "grad_norm": 0.5731651782989502,
      "learning_rate": 8.82706015034849e-06,
      "loss": 0.0622,
      "step": 260
    },
    {
      "epoch": 4.993166287015946,
      "eval_loss": 0.09794322401285172,
      "eval_runtime": 48.874,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 274
    },
    {
      "epoch": 5.10250569476082,
      "grad_norm": 0.5230545401573181,
      "learning_rate": 8.527703282684766e-06,
      "loss": 0.062,
      "step": 280
    },
    {
      "epoch": 5.466970387243736,
      "grad_norm": 0.631493330001831,
      "learning_rate": 8.20073735379715e-06,
      "loss": 0.0577,
      "step": 300
    },
    {
      "epoch": 5.831435079726651,
      "grad_norm": 0.7749846577644348,
      "learning_rate": 7.848721315395975e-06,
      "loss": 0.0576,
      "step": 320
    },
    {
      "epoch": 5.995444191343964,
      "eval_loss": 0.09521909803152084,
      "eval_runtime": 48.8686,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.43,
      "step": 329
    }
  ],
  "logging_steps": 20,
  "max_steps": 810,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 1.2319125158965248e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}