trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.999812605409457,
  "global_step": 6003,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 1.1642,
      "step": 500
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 0.9403,
      "step": 1000
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 0.8969,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 0.871,
      "step": 2000
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 0.5011,
      "step": 2500
    },
    {
      "epoch": 1.5,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 3000
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "learning_rate": 2e-05,
      "loss": 0.528,
      "step": 4000
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 0.2731,
      "step": 4500
    },
    {
      "epoch": 2.5,
      "learning_rate": 2e-05,
      "loss": 0.2872,
      "step": 5000
    },
    {
      "epoch": 2.75,
      "learning_rate": 2e-05,
      "loss": 0.2964,
      "step": 5500
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-05,
      "loss": 0.3057,
      "step": 6000
    },
    {
      "epoch": 3.0,
      "step": 6003,
      "total_flos": 48405780234240.0,
      "train_loss": 0.5916195661747989,
      "train_runtime": 62605.4207,
      "train_samples_per_second": 6.137,
      "train_steps_per_second": 0.096
    }
  ],
  "max_steps": 6003,
  "num_train_epochs": 3,
  "total_flos": 48405780234240.0,
  "trial_name": null,
  "trial_params": null
}