trainer_state.json

{
  "best_metric": 0.8568,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0772/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": 91.47213745117188,
      "learning_rate": 6.653391037658466e-05,
      "loss": 3.2311,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6677333333333333,
      "eval_loss": 1.511957049369812,
      "eval_runtime": 7.732,
      "eval_samples_per_second": 484.999,
      "eval_steps_per_second": 7.631,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 89.09217834472656,
      "learning_rate": 5.682214306505567e-05,
      "loss": 0.9173,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8093333333333333,
      "eval_loss": 0.6653686165809631,
      "eval_runtime": 7.6236,
      "eval_samples_per_second": 491.896,
      "eval_steps_per_second": 7.739,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 78.70437622070312,
      "learning_rate": 4.2788232688471e-05,
      "loss": 0.5274,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8394666666666667,
      "eval_loss": 0.5181002616882324,
      "eval_runtime": 8.0843,
      "eval_samples_per_second": 463.86,
      "eval_steps_per_second": 7.298,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 65.95828247070312,
      "learning_rate": 2.7211767311528996e-05,
      "loss": 0.378,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8517333333333333,
      "eval_loss": 0.4700625240802765,
      "eval_runtime": 7.6472,
      "eval_samples_per_second": 490.376,
      "eval_steps_per_second": 7.715,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 141.72393798828125,
      "learning_rate": 1.3177856934944328e-05,
      "loss": 0.3017,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8549333333333333,
      "eval_loss": 0.4613685607910156,
      "eval_runtime": 7.631,
      "eval_samples_per_second": 491.418,
      "eval_steps_per_second": 7.732,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 75.83885192871094,
      "learning_rate": 3.4660896234153336e-06,
      "loss": 0.2592,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.45299890637397766,
      "eval_runtime": 7.6544,
      "eval_samples_per_second": 489.915,
      "eval_steps_per_second": 7.708,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 70.49552917480469,
      "learning_rate": 0.0,
      "loss": 0.2486,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.4609434902667999,
      "eval_runtime": 7.6915,
      "eval_samples_per_second": 487.55,
      "eval_steps_per_second": 7.671,
      "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
}