{
  "best_metric": 0.747850775718689,
  "best_model_checkpoint": "Action_model/checkpoint-300",
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 314,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 1.3472708463668823,
      "learning_rate": 9.363057324840766e-05,
      "loss": 2.1856,
      "step": 20
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6540354490280151,
      "learning_rate": 8.726114649681529e-05,
      "loss": 1.9121,
      "step": 40
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.9076273441314697,
      "learning_rate": 8.089171974522294e-05,
      "loss": 1.5695,
      "step": 60
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.171828508377075,
      "learning_rate": 7.452229299363057e-05,
      "loss": 1.3273,
      "step": 80
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.170124053955078,
      "learning_rate": 6.815286624203822e-05,
      "loss": 1.1778,
      "step": 100
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.78,
      "eval_loss": 1.0651038885116577,
      "eval_runtime": 20.8448,
      "eval_samples_per_second": 50.372,
      "eval_steps_per_second": 6.333,
      "step": 100
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.705270767211914,
      "learning_rate": 6.178343949044585e-05,
      "loss": 1.0757,
      "step": 120
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.980999231338501,
      "learning_rate": 5.5414012738853505e-05,
      "loss": 1.0115,
      "step": 140
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.483489990234375,
      "learning_rate": 4.904458598726115e-05,
      "loss": 0.9806,
      "step": 160
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.3504586219787598,
      "learning_rate": 4.267515923566879e-05,
      "loss": 0.8805,
      "step": 180
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.248554229736328,
      "learning_rate": 3.630573248407643e-05,
      "loss": 0.8527,
      "step": 200
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.8133333333333334,
      "eval_loss": 0.7873561978340149,
      "eval_runtime": 14.6757,
      "eval_samples_per_second": 71.547,
      "eval_steps_per_second": 8.994,
      "step": 200
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.6971187591552734,
      "learning_rate": 2.9936305732484078e-05,
      "loss": 0.7773,
      "step": 220
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.3084142208099365,
      "learning_rate": 2.356687898089172e-05,
      "loss": 0.6965,
      "step": 240
    },
    {
      "epoch": 0.83,
      "grad_norm": 3.3731772899627686,
      "learning_rate": 1.7197452229299362e-05,
      "loss": 0.77,
      "step": 260
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8178930282592773,
      "learning_rate": 1.0828025477707008e-05,
      "loss": 0.7106,
      "step": 280
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.819934368133545,
      "learning_rate": 4.45859872611465e-06,
      "loss": 0.7114,
      "step": 300
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.8095238095238095,
      "eval_loss": 0.747850775718689,
      "eval_runtime": 14.738,
      "eval_samples_per_second": 71.244,
      "eval_steps_per_second": 8.956,
      "step": 300
    },
    {
      "epoch": 1.0,
      "step": 314,
      "total_flos": 7.776878731479245e+17,
      "train_loss": 1.0910153146002703,
      "train_runtime": 357.1216,
      "train_samples_per_second": 28.1,
      "train_steps_per_second": 0.879
    }
  ],
  "logging_steps": 20,
  "max_steps": 314,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 7.776878731479245e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}