trainer_state.json

{
  "best_metric": 0.8658666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0550/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 171.46324157714844,
      "learning_rate": 6.653391037658466e-05,
      "loss": 3.262,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6536,
      "eval_loss": 1.7230042219161987,
      "eval_runtime": 7.3009,
      "eval_samples_per_second": 513.633,
      "eval_steps_per_second": 8.081,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 192.272216796875,
      "learning_rate": 5.682214306505567e-05,
      "loss": 1.0023,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8114666666666667,
      "eval_loss": 0.6556878089904785,
      "eval_runtime": 7.1855,
      "eval_samples_per_second": 521.885,
      "eval_steps_per_second": 8.211,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 174.25746154785156,
      "learning_rate": 4.2788232688471e-05,
      "loss": 0.529,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8418666666666667,
      "eval_loss": 0.5055389404296875,
      "eval_runtime": 7.2884,
      "eval_samples_per_second": 514.517,
      "eval_steps_per_second": 8.095,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 162.2234649658203,
      "learning_rate": 2.7211767311528996e-05,
      "loss": 0.3759,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8549333333333333,
      "eval_loss": 0.46235138177871704,
      "eval_runtime": 7.2259,
      "eval_samples_per_second": 518.963,
      "eval_steps_per_second": 8.165,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 240.05418395996094,
      "learning_rate": 1.3177856934944328e-05,
      "loss": 0.2975,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.45367851853370667,
      "eval_runtime": 7.1445,
      "eval_samples_per_second": 524.882,
      "eval_steps_per_second": 8.258,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 143.2042694091797,
      "learning_rate": 3.4660896234153336e-06,
      "loss": 0.2597,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.43605566024780273,
      "eval_runtime": 7.1359,
      "eval_samples_per_second": 525.512,
      "eval_steps_per_second": 8.268,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 145.73959350585938,
      "learning_rate": 0.0,
      "loss": 0.2462,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.4272597134113312,
      "eval_runtime": 7.3654,
      "eval_samples_per_second": 509.139,
      "eval_steps_per_second": 8.01,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}