trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "global_step": 7470,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 4.6653279785809906e-05,
      "loss": 2.185,
      "step": 500
    },
    {
      "epoch": 2.01,
      "learning_rate": 4.3306559571619816e-05,
      "loss": 1.8076,
      "step": 1000
    },
    {
      "epoch": 3.01,
      "learning_rate": 3.995983935742972e-05,
      "loss": 1.5817,
      "step": 1500
    },
    {
      "epoch": 4.02,
      "learning_rate": 3.661311914323963e-05,
      "loss": 1.3979,
      "step": 2000
    },
    {
      "epoch": 5.02,
      "learning_rate": 3.326639892904953e-05,
      "loss": 1.2541,
      "step": 2500
    },
    {
      "epoch": 6.02,
      "learning_rate": 2.991967871485944e-05,
      "loss": 1.127,
      "step": 3000
    },
    {
      "epoch": 7.03,
      "learning_rate": 2.6572958500669343e-05,
      "loss": 1.0202,
      "step": 3500
    },
    {
      "epoch": 8.03,
      "learning_rate": 2.322623828647925e-05,
      "loss": 0.9236,
      "step": 4000
    },
    {
      "epoch": 9.04,
      "learning_rate": 1.9879518072289157e-05,
      "loss": 0.8492,
      "step": 4500
    },
    {
      "epoch": 10.04,
      "learning_rate": 1.6532797858099064e-05,
      "loss": 0.7856,
      "step": 5000
    },
    {
      "epoch": 11.04,
      "learning_rate": 1.318607764390897e-05,
      "loss": 0.7317,
      "step": 5500
    },
    {
      "epoch": 12.05,
      "learning_rate": 9.839357429718876e-06,
      "loss": 0.6839,
      "step": 6000
    },
    {
      "epoch": 13.05,
      "learning_rate": 6.492637215528783e-06,
      "loss": 0.6487,
      "step": 6500
    },
    {
      "epoch": 14.06,
      "learning_rate": 3.1459170013386886e-06,
      "loss": 0.6207,
      "step": 7000
    },
    {
      "epoch": 15.0,
      "step": 7470,
      "total_flos": 2.979518169858048e+16,
      "train_loss": 1.0831103017212078,
      "train_runtime": 1681.1253,
      "train_samples_per_second": 142.146,
      "train_steps_per_second": 4.443
    }
  ],
  "max_steps": 7470,
  "num_train_epochs": 15,
  "total_flos": 2.979518169858048e+16,
  "trial_name": null,
  "trial_params": null
}