trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5268,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2847380410022779,
      "grad_norm": 2.7607381343841553,
      "learning_rate": 4.525436598329537e-05,
      "loss": 0.1616,
      "step": 500
    },
    {
      "epoch": 0.5694760820045558,
      "grad_norm": 0.94535231590271,
      "learning_rate": 4.050873196659074e-05,
      "loss": 0.0728,
      "step": 1000
    },
    {
      "epoch": 0.8542141230068337,
      "grad_norm": 2.5376665592193604,
      "learning_rate": 3.5763097949886106e-05,
      "loss": 0.063,
      "step": 1500
    },
    {
      "epoch": 1.1389521640091116,
      "grad_norm": 1.2084846496582031,
      "learning_rate": 3.1017463933181475e-05,
      "loss": 0.0425,
      "step": 2000
    },
    {
      "epoch": 1.4236902050113895,
      "grad_norm": 0.5610941648483276,
      "learning_rate": 2.6271829916476843e-05,
      "loss": 0.0298,
      "step": 2500
    },
    {
      "epoch": 1.7084282460136673,
      "grad_norm": 0.018142463639378548,
      "learning_rate": 2.152619589977221e-05,
      "loss": 0.027,
      "step": 3000
    },
    {
      "epoch": 1.9931662870159452,
      "grad_norm": 3.1458160877227783,
      "learning_rate": 1.678056188306758e-05,
      "loss": 0.0248,
      "step": 3500
    },
    {
      "epoch": 2.277904328018223,
      "grad_norm": 0.005340315867215395,
      "learning_rate": 1.2034927866362947e-05,
      "loss": 0.0116,
      "step": 4000
    },
    {
      "epoch": 2.562642369020501,
      "grad_norm": 0.048937492072582245,
      "learning_rate": 7.289293849658315e-06,
      "loss": 0.0116,
      "step": 4500
    },
    {
      "epoch": 2.847380410022779,
      "grad_norm": 0.005267954897135496,
      "learning_rate": 2.5436598329536827e-06,
      "loss": 0.0093,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "step": 5268,
      "total_flos": 891900038010780.0,
      "train_loss": 0.043606681179258075,
      "train_runtime": 849.8761,
      "train_samples_per_second": 49.564,
      "train_steps_per_second": 6.199
    }
  ],
  "logging_steps": 500,
  "max_steps": 5268,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 891900038010780.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}