trainer_state.json

{
  "best_metric": 0.84,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0940/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": 278.6871643066406,
      "learning_rate": 8.535533905932738e-05,
      "loss": 2.8135,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6944,
      "eval_loss": 1.166214108467102,
      "eval_runtime": 8.7717,
      "eval_samples_per_second": 427.512,
      "eval_steps_per_second": 6.726,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 86.42532348632812,
      "learning_rate": 5e-05,
      "loss": 0.8409,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8082666666666667,
      "eval_loss": 0.6608612537384033,
      "eval_runtime": 8.5419,
      "eval_samples_per_second": 439.011,
      "eval_steps_per_second": 6.907,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 241.2579803466797,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.5626,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5537157654762268,
      "eval_runtime": 9.0862,
      "eval_samples_per_second": 412.712,
      "eval_steps_per_second": 6.493,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 1332,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}