trainer_state.json

{
  "best_metric": 0.8901333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0887/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": 115.39852905273438,
      "learning_rate": 8.75e-05,
      "loss": 2.6958,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7424,
      "eval_loss": 0.9329473376274109,
      "eval_runtime": 8.116,
      "eval_samples_per_second": 462.053,
      "eval_steps_per_second": 7.27,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 166.97422790527344,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.6981,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5381203293800354,
      "eval_runtime": 8.0083,
      "eval_samples_per_second": 468.264,
      "eval_steps_per_second": 7.367,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 64.7828598022461,
      "learning_rate": 6.25e-05,
      "loss": 0.4118,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.4497877359390259,
      "eval_runtime": 8.0456,
      "eval_samples_per_second": 466.091,
      "eval_steps_per_second": 7.333,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 108.96895599365234,
      "learning_rate": 5e-05,
      "loss": 0.276,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.4111157953739166,
      "eval_runtime": 9.394,
      "eval_samples_per_second": 399.192,
      "eval_steps_per_second": 6.281,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 101.89673614501953,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.1943,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8861333333333333,
      "eval_loss": 0.392296701669693,
      "eval_runtime": 8.0356,
      "eval_samples_per_second": 466.671,
      "eval_steps_per_second": 7.342,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 80.19306945800781,
      "learning_rate": 2.5e-05,
      "loss": 0.1483,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.38535526394844055,
      "eval_runtime": 7.8902,
      "eval_samples_per_second": 475.274,
      "eval_steps_per_second": 7.478,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 192.78701782226562,
      "learning_rate": 1.25e-05,
      "loss": 0.1195,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8901333333333333,
      "eval_loss": 0.37588655948638916,
      "eval_runtime": 8.0691,
      "eval_samples_per_second": 464.734,
      "eval_steps_per_second": 7.312,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}