trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 6402,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "learning_rate": 4.609497032177445e-05,
      "loss": 3.0016,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.218994064354889e-05,
      "loss": 2.6439,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.828491096532334e-05,
      "loss": 2.4277,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.437988128709778e-05,
      "loss": 2.27,
      "step": 2000
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.047485160887223e-05,
      "loss": 1.9681,
      "step": 2500
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.6569821930646678e-05,
      "loss": 1.8284,
      "step": 3000
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.266479225242112e-05,
      "loss": 1.7601,
      "step": 3500
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.8759762574195563e-05,
      "loss": 1.7051,
      "step": 4000
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4854732895970011e-05,
      "loss": 1.595,
      "step": 4500
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.0949703217744455e-05,
      "loss": 1.495,
      "step": 5000
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.0446735395189e-06,
      "loss": 1.4778,
      "step": 5500
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.1396438612933463e-06,
      "loss": 1.4717,
      "step": 6000
    },
    {
      "epoch": 3.0,
      "step": 6402,
      "total_flos": 438186742382592.0,
      "train_runtime": 3295.7162,
      "train_samples_per_second": 1.943
    }
  ],
  "max_steps": 6402,
  "num_train_epochs": 3,
  "total_flos": 438186742382592.0,
  "trial_name": null,
  "trial_params": null
}