last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10227272727272728,
  "eval_steps": 9,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011363636363636364,
      "eval_loss": 3.371885061264038,
      "eval_runtime": 16.7352,
      "eval_samples_per_second": 8.844,
      "eval_steps_per_second": 1.135,
      "step": 1
    },
    {
      "epoch": 0.03409090909090909,
      "grad_norm": 0.9526873826980591,
      "learning_rate": 1.5e-05,
      "loss": 3.4877,
      "step": 3
    },
    {
      "epoch": 0.06818181818181818,
      "grad_norm": 0.9885739088058472,
      "learning_rate": 3e-05,
      "loss": 3.3668,
      "step": 6
    },
    {
      "epoch": 0.10227272727272728,
      "grad_norm": 1.137595772743225,
      "learning_rate": 4.5e-05,
      "loss": 3.5998,
      "step": 9
    },
    {
      "epoch": 0.10227272727272728,
      "eval_loss": 3.313127279281616,
      "eval_runtime": 17.0437,
      "eval_samples_per_second": 8.684,
      "eval_steps_per_second": 1.115,
      "step": 9
    }
  ],
  "logging_steps": 3,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 9,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.0621641453142016e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}