trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.005737563830397613,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 1.2234,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00019396926207859084,
      "loss": 1.1179,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.0056,
      "step": 40
    },
    {
      "epoch": 0.0,
      "learning_rate": 8.263518223330697e-05,
      "loss": 0.9795,
      "step": 60
    },
    {
      "epoch": 0.0,
      "learning_rate": 2.339555568810221e-05,
      "loss": 0.9954,
      "step": 80
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0,
      "loss": 0.9545,
      "step": 100
    },
    {
      "epoch": 0.01,
      "step": 100,
      "total_flos": 7.072526874181632e+16,
      "train_loss": 1.0116421318054198,
      "train_runtime": 473.1338,
      "train_samples_per_second": 1.691,
      "train_steps_per_second": 0.211
    }
  ],
  "logging_steps": 20,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 7.072526874181632e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}