trainer_state.json

{
  "best_metric": 0.7477333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0140/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": 149.2130889892578,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 3.8588,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3232,
      "eval_loss": 3.7513582706451416,
      "eval_runtime": 15.5421,
      "eval_samples_per_second": 241.281,
      "eval_steps_per_second": 3.796,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 201.67662048339844,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 3.4282,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5261333333333333,
      "eval_loss": 3.037914276123047,
      "eval_runtime": 15.8028,
      "eval_samples_per_second": 237.3,
      "eval_steps_per_second": 3.734,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 361.1563720703125,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 2.4707,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6261333333333333,
      "eval_loss": 1.932478904724121,
      "eval_runtime": 15.382,
      "eval_samples_per_second": 243.791,
      "eval_steps_per_second": 3.836,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 315.8300476074219,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 1.6643,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6858666666666666,
      "eval_loss": 1.4152746200561523,
      "eval_runtime": 15.3089,
      "eval_samples_per_second": 244.955,
      "eval_steps_per_second": 3.854,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 513.2781372070312,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.323,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7144,
      "eval_loss": 1.1795281171798706,
      "eval_runtime": 15.3786,
      "eval_samples_per_second": 243.845,
      "eval_steps_per_second": 3.836,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 314.75067138671875,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.1582,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7410666666666667,
      "eval_loss": 1.068315863609314,
      "eval_runtime": 15.4033,
      "eval_samples_per_second": 243.455,
      "eval_steps_per_second": 3.83,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 366.9861145019531,
      "learning_rate": 0.0,
      "loss": 1.0854,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7477333333333334,
      "eval_loss": 1.0448601245880127,
      "eval_runtime": 15.4282,
      "eval_samples_per_second": 243.061,
      "eval_steps_per_second": 3.824,
      "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
}