trainer_state.json

{
  "best_metric": 0.8821333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0763/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": 76.3219223022461,
      "learning_rate": 5e-05,
      "loss": 3.6636,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5213333333333333,
      "eval_loss": 2.953976631164551,
      "eval_runtime": 13.4455,
      "eval_samples_per_second": 278.904,
      "eval_steps_per_second": 4.388,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 269.1883239746094,
      "learning_rate": 5e-05,
      "loss": 1.679,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7690666666666667,
      "eval_loss": 0.9018044471740723,
      "eval_runtime": 13.0125,
      "eval_samples_per_second": 288.185,
      "eval_steps_per_second": 4.534,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 303.69964599609375,
      "learning_rate": 5e-05,
      "loss": 0.7648,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8282666666666667,
      "eval_loss": 0.6057847738265991,
      "eval_runtime": 13.1162,
      "eval_samples_per_second": 285.905,
      "eval_steps_per_second": 4.498,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 100.72391510009766,
      "learning_rate": 5e-05,
      "loss": 0.5445,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8512,
      "eval_loss": 0.49447882175445557,
      "eval_runtime": 12.9655,
      "eval_samples_per_second": 289.229,
      "eval_steps_per_second": 4.551,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 95.46270751953125,
      "learning_rate": 5e-05,
      "loss": 0.4295,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.4192505478858948,
      "eval_runtime": 12.9383,
      "eval_samples_per_second": 289.837,
      "eval_steps_per_second": 4.56,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 166.97531127929688,
      "learning_rate": 5e-05,
      "loss": 0.3451,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.40309545397758484,
      "eval_runtime": 13.2625,
      "eval_samples_per_second": 282.751,
      "eval_steps_per_second": 4.449,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 246.95433044433594,
      "learning_rate": 5e-05,
      "loss": 0.2953,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8821333333333333,
      "eval_loss": 0.38451313972473145,
      "eval_runtime": 12.7789,
      "eval_samples_per_second": 293.452,
      "eval_steps_per_second": 4.617,
      "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
}