last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10607734806629834,
  "eval_steps": 9,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011786372007366482,
      "eval_loss": 3.2576773166656494,
      "eval_runtime": 48.3038,
      "eval_samples_per_second": 5.921,
      "eval_steps_per_second": 2.96,
      "step": 1
    },
    {
      "epoch": 0.03535911602209945,
      "grad_norm": 37.830448150634766,
      "learning_rate": 2.2499999999999998e-05,
      "loss": 99.4918,
      "step": 3
    },
    {
      "epoch": 0.0707182320441989,
      "grad_norm": 30.299089431762695,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 95.1039,
      "step": 6
    },
    {
      "epoch": 0.10607734806629834,
      "grad_norm": 31.264686584472656,
      "learning_rate": 6.75e-05,
      "loss": 99.6355,
      "step": 9
    },
    {
      "epoch": 0.10607734806629834,
      "eval_loss": 2.994905948638916,
      "eval_runtime": 48.3217,
      "eval_samples_per_second": 5.919,
      "eval_steps_per_second": 2.959,
      "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": 4.733240683895194e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}