trainer_state.json

{
  "best_metric": 5474.208984375,
  "best_model_checkpoint": "./coco_outputs/checkpoint-20",
  "epoch": 5.0,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 5475.6171875,
      "eval_runtime": 2.202,
      "eval_samples_per_second": 14.532,
      "eval_steps_per_second": 1.816,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_loss": 5475.1103515625,
      "eval_runtime": 2.2765,
      "eval_samples_per_second": 14.056,
      "eval_steps_per_second": 1.757,
      "step": 8
    },
    {
      "epoch": 2.5,
      "learning_rate": 1e-05,
      "loss": 4957.6352,
      "step": 10
    },
    {
      "epoch": 3.0,
      "eval_loss": 5474.685546875,
      "eval_runtime": 2.2499,
      "eval_samples_per_second": 14.223,
      "eval_steps_per_second": 1.778,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_loss": 5474.34375,
      "eval_runtime": 2.2284,
      "eval_samples_per_second": 14.36,
      "eval_steps_per_second": 1.795,
      "step": 16
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0,
      "loss": 5016.0344,
      "step": 20
    },
    {
      "epoch": 5.0,
      "eval_loss": 5474.208984375,
      "eval_runtime": 2.3127,
      "eval_samples_per_second": 13.837,
      "eval_steps_per_second": 1.73,
      "step": 20
    },
    {
      "epoch": 5.0,
      "step": 20,
      "total_flos": 7.65002115072e+16,
      "train_loss": 4986.834765625,
      "train_runtime": 89.7426,
      "train_samples_per_second": 1.783,
      "train_steps_per_second": 0.223
    }
  ],
  "max_steps": 20,
  "num_train_epochs": 5,
  "total_flos": 7.65002115072e+16,
  "trial_name": null,
  "trial_params": null
}