trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9982748706152962,
  "global_step": 434,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "learning_rate": 3.7500000000000005e-08,
      "loss": 1.2688,
      "step": 100
    },
    {
      "epoch": 0.46,
      "learning_rate": 7.500000000000001e-08,
      "loss": 1.2572,
      "step": 200
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.125e-07,
      "loss": 1.2306,
      "step": 300
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 1.2477,
      "step": 400
    },
    {
      "epoch": 1.0,
      "step": 434,
      "total_flos": 1.2564599871773491e+17,
      "train_loss": 1.2537303696030295,
      "train_runtime": 320.3475,
      "train_samples_per_second": 10.857,
      "train_steps_per_second": 1.355
    }
  ],
  "max_steps": 434,
  "num_train_epochs": 1,
  "total_flos": 1.2564599871773491e+17,
  "trial_name": null,
  "trial_params": null
}