trainer_state.json

{
  "best_metric": 0.5781333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0300/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 132.040771484375,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 3.8532,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3464,
      "eval_loss": 3.745779275894165,
      "eval_runtime": 7.4962,
      "eval_samples_per_second": 500.251,
      "eval_steps_per_second": 7.871,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 197.33290100097656,
      "learning_rate": 9.999999999999999e-06,
      "loss": 3.5369,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5344,
      "eval_loss": 3.314164400100708,
      "eval_runtime": 7.3274,
      "eval_samples_per_second": 511.778,
      "eval_steps_per_second": 8.052,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 359.0583190917969,
      "learning_rate": 0.0,
      "loss": 3.168,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5781333333333334,
      "eval_loss": 3.1841394901275635,
      "eval_runtime": 7.4351,
      "eval_samples_per_second": 504.364,
      "eval_steps_per_second": 7.935,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "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": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}