trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.966292134831461,
  "eval_steps": 500,
  "global_step": 33,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.898876404494382,
      "grad_norm": 3.078231865700168,
      "learning_rate": 5e-06,
      "loss": 1.0475,
      "step": 10
    },
    {
      "epoch": 0.9887640449438202,
      "eval_loss": 0.9338513612747192,
      "eval_runtime": 12.6708,
      "eval_samples_per_second": 23.519,
      "eval_steps_per_second": 0.395,
      "step": 11
    },
    {
      "epoch": 1.797752808988764,
      "grad_norm": 1.4213980745912356,
      "learning_rate": 5e-06,
      "loss": 0.9997,
      "step": 20
    },
    {
      "epoch": 1.9775280898876404,
      "eval_loss": 0.8875383138656616,
      "eval_runtime": 12.5779,
      "eval_samples_per_second": 23.692,
      "eval_steps_per_second": 0.398,
      "step": 22
    },
    {
      "epoch": 2.696629213483146,
      "grad_norm": 1.4804640058828842,
      "learning_rate": 5e-06,
      "loss": 0.9271,
      "step": 30
    },
    {
      "epoch": 2.966292134831461,
      "eval_loss": 0.8657615780830383,
      "eval_runtime": 11.9918,
      "eval_samples_per_second": 24.85,
      "eval_steps_per_second": 0.417,
      "step": 33
    },
    {
      "epoch": 2.966292134831461,
      "step": 33,
      "total_flos": 55066849443840.0,
      "train_loss": 0.9766716306859796,
      "train_runtime": 2228.2703,
      "train_samples_per_second": 7.601,
      "train_steps_per_second": 0.015
    }
  ],
  "logging_steps": 10,
  "max_steps": 33,
  "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": 55066849443840.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}