trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 111,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 8.75682357849744,
      "learning_rate": 5e-06,
      "loss": 0.9901,
      "step": 10
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 2.499156992011751,
      "learning_rate": 5e-06,
      "loss": 0.8867,
      "step": 20
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.0566032870943596,
      "learning_rate": 5e-06,
      "loss": 0.8501,
      "step": 30
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 2.007752272774514,
      "learning_rate": 5e-06,
      "loss": 0.8236,
      "step": 40
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 0.9608340664542014,
      "learning_rate": 5e-06,
      "loss": 0.7835,
      "step": 50
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.0118488800156256,
      "learning_rate": 5e-06,
      "loss": 0.7712,
      "step": 60
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.6936595596230254,
      "learning_rate": 5e-06,
      "loss": 0.7645,
      "step": 70
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.8650689649871667,
      "learning_rate": 5e-06,
      "loss": 0.736,
      "step": 80
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 0.6831306747324589,
      "learning_rate": 5e-06,
      "loss": 0.7059,
      "step": 90
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.5685808688521499,
      "learning_rate": 5e-06,
      "loss": 0.7043,
      "step": 100
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 1.415202193366322,
      "learning_rate": 5e-06,
      "loss": 0.7053,
      "step": 110
    },
    {
      "epoch": 3.0,
      "step": 111,
      "total_flos": 185510374932480.0,
      "train_loss": 0.7921632098722028,
      "train_runtime": 1731.9034,
      "train_samples_per_second": 32.782,
      "train_steps_per_second": 0.064
    }
  ],
  "logging_steps": 10,
  "max_steps": 111,
  "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": 185510374932480.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}