{
  "best_metric": 0.10630220197418375,
  "best_model_checkpoint": "microsoft/beit-proposed-1-lr-0.001/checkpoint-988",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 988,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "grad_norm": 2.142606735229492,
      "learning_rate": 0.0009988751406074241,
      "loss": 6.1812,
      "step": 100
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.128974437713623,
      "learning_rate": 0.0008863892013498312,
      "loss": 5.826,
      "step": 200
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6932368278503418,
      "learning_rate": 0.0007739032620922385,
      "loss": 5.5885,
      "step": 300
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.06595516204834,
      "learning_rate": 0.0006614173228346457,
      "loss": 5.4277,
      "step": 400
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.4056410789489746,
      "learning_rate": 0.0005489313835770528,
      "loss": 5.1868,
      "step": 500
    },
    {
      "epoch": 0.61,
      "grad_norm": 3.8991808891296387,
      "learning_rate": 0.0004364454443194601,
      "loss": 4.9279,
      "step": 600
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.9080541133880615,
      "learning_rate": 0.0003239595050618673,
      "loss": 4.7066,
      "step": 700
    },
    {
      "epoch": 0.81,
      "grad_norm": 3.6365692615509033,
      "learning_rate": 0.00021147356580427447,
      "loss": 4.5313,
      "step": 800
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.200676441192627,
      "learning_rate": 9.898762654668166e-05,
      "loss": 4.3665,
      "step": 900
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.10630220197418375,
      "eval_loss": 4.188309669494629,
      "eval_runtime": 22.9496,
      "eval_samples_per_second": 172.16,
      "eval_steps_per_second": 5.403,
      "step": 988
    },
    {
      "epoch": 1.0,
      "step": 988,
      "total_flos": 2.457913337181389e+18,
      "train_loss": 5.110672645723289,
      "train_runtime": 357.0386,
      "train_samples_per_second": 88.528,
      "train_steps_per_second": 2.767
    }
  ],
  "logging_steps": 100,
  "max_steps": 988,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 2.457913337181389e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}