trainer_state.json

{
  "best_metric": 0.8837333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0983/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": 86.88298797607422,
      "learning_rate": 0.000225,
      "loss": 1.677,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8186666666666667,
      "eval_loss": 0.5815926790237427,
      "eval_runtime": 9.421,
      "eval_samples_per_second": 398.048,
      "eval_steps_per_second": 6.263,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 44.34465408325195,
      "learning_rate": 0.00015,
      "loss": 0.4622,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.44064581394195557,
      "eval_runtime": 7.4013,
      "eval_samples_per_second": 506.668,
      "eval_steps_per_second": 7.972,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 37.54586410522461,
      "learning_rate": 7.5e-05,
      "loss": 0.2661,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8837333333333334,
      "eval_loss": 0.38763293623924255,
      "eval_runtime": 7.427,
      "eval_samples_per_second": 504.914,
      "eval_steps_per_second": 7.944,
      "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
}