trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 9564,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 0.0001,
      "loss": 2.2922,
      "step": 500
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001,
      "loss": 2.1835,
      "step": 1000
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001,
      "loss": 2.1451,
      "step": 1500
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0001,
      "loss": 2.1083,
      "step": 2000
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0001,
      "loss": 2.0944,
      "step": 2500
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0001,
      "loss": 2.0719,
      "step": 3000
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0001,
      "loss": 2.0304,
      "step": 3500
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0001,
      "loss": 1.9961,
      "step": 4000
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0001,
      "loss": 1.9908,
      "step": 4500
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0001,
      "loss": 1.9804,
      "step": 5000
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0001,
      "loss": 1.9774,
      "step": 5500
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0001,
      "loss": 1.9687,
      "step": 6000
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0001,
      "loss": 1.9451,
      "step": 6500
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0001,
      "loss": 1.902,
      "step": 7000
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0001,
      "loss": 1.9051,
      "step": 7500
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.0001,
      "loss": 1.8918,
      "step": 8000
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0001,
      "loss": 1.8936,
      "step": 8500
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.0001,
      "loss": 1.8971,
      "step": 9000
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0001,
      "loss": 1.8882,
      "step": 9500
    },
    {
      "epoch": 3.0,
      "step": 9564,
      "total_flos": 7.159992935875707e+17,
      "train_loss": 2.007772041234647,
      "train_runtime": 34797.0582,
      "train_samples_per_second": 17.589,
      "train_steps_per_second": 0.275
    }
  ],
  "max_steps": 9564,
  "num_train_epochs": 3,
  "total_flos": 7.159992935875707e+17,
  "trial_name": null,
  "trial_params": null
}