trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1205,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 5302.2344,
      "step": 100
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 2147.5883,
      "step": 200
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 2079.1225,
      "step": 300
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 924.7552,
      "step": 400
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 468.9583,
      "step": 500
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0002,
      "loss": 365.4921,
      "step": 600
    },
    {
      "epoch": 2.9,
      "learning_rate": 0.0002,
      "loss": 293.9367,
      "step": 700
    },
    {
      "epoch": 3.32,
      "learning_rate": 0.0002,
      "loss": 236.9112,
      "step": 800
    },
    {
      "epoch": 3.73,
      "learning_rate": 0.0002,
      "loss": 211.7118,
      "step": 900
    },
    {
      "epoch": 4.15,
      "learning_rate": 0.0002,
      "loss": 159.076,
      "step": 1000
    },
    {
      "epoch": 4.56,
      "learning_rate": 0.0002,
      "loss": 149.862,
      "step": 1100
    },
    {
      "epoch": 4.98,
      "learning_rate": 0.0002,
      "loss": 121.4963,
      "step": 1200
    },
    {
      "epoch": 5.0,
      "step": 1205,
      "total_flos": 3.9380753570781855e+18,
      "train_loss": 1034.4726613658097,
      "train_runtime": 1395.2083,
      "train_samples_per_second": 13.779,
      "train_steps_per_second": 0.864
    }
  ],
  "logging_steps": 100,
  "max_steps": 1205,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 3.9380753570781855e+18,
  "trial_name": null,
  "trial_params": null
}