trainer_state.json

{
  "best_metric": 0.8741333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0607/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": 68.32880401611328,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.6819,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7581333333333333,
      "eval_loss": 0.8816426992416382,
      "eval_runtime": 9.4684,
      "eval_samples_per_second": 396.056,
      "eval_steps_per_second": 6.231,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 200.93524169921875,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.652,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8298666666666666,
      "eval_loss": 0.52366042137146,
      "eval_runtime": 9.1968,
      "eval_samples_per_second": 407.752,
      "eval_steps_per_second": 6.415,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 110.75025177001953,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.3979,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.46227553486824036,
      "eval_runtime": 8.9874,
      "eval_samples_per_second": 417.25,
      "eval_steps_per_second": 6.565,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 66.47645568847656,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.2786,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8624,
      "eval_loss": 0.4232029914855957,
      "eval_runtime": 8.9648,
      "eval_samples_per_second": 418.301,
      "eval_steps_per_second": 6.581,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 84.39371490478516,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.2103,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.41237759590148926,
      "eval_runtime": 8.9259,
      "eval_samples_per_second": 420.125,
      "eval_steps_per_second": 6.61,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 77.57479095458984,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.1671,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.41748130321502686,
      "eval_runtime": 8.816,
      "eval_samples_per_second": 425.365,
      "eval_steps_per_second": 6.692,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 64.06316375732422,
      "learning_rate": 0.0,
      "loss": 0.1417,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8741333333333333,
      "eval_loss": 0.4093964695930481,
      "eval_runtime": 8.8369,
      "eval_samples_per_second": 424.355,
      "eval_steps_per_second": 6.677,
      "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
}