trainer_state.json

{
  "best_metric": 0.9034666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0618/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": Infinity,
      "learning_rate": 0.00047538820348083834,
      "loss": 0.8648,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7989333333333334,
      "eval_loss": 0.688944399356842,
      "eval_runtime": 11.0794,
      "eval_samples_per_second": 338.466,
      "eval_steps_per_second": 5.325,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8822070956230164,
      "learning_rate": 0.00040613573610928476,
      "loss": 0.4786,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8208,
      "eval_loss": 0.6280103325843811,
      "eval_runtime": 11.0375,
      "eval_samples_per_second": 339.751,
      "eval_steps_per_second": 5.345,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.12453437596559525,
      "learning_rate": 0.0003059586713221479,
      "loss": 0.2873,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8514666666666667,
      "eval_loss": 0.5088369846343994,
      "eval_runtime": 11.1412,
      "eval_samples_per_second": 336.588,
      "eval_steps_per_second": 5.296,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.07038894295692444,
      "learning_rate": 0.00019469830539159362,
      "loss": 0.1529,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.45563164353370667,
      "eval_runtime": 11.1661,
      "eval_samples_per_second": 335.837,
      "eval_steps_per_second": 5.284,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.005879338830709457,
      "learning_rate": 9.43911183090096e-05,
      "loss": 0.0484,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.44927480816841125,
      "eval_runtime": 11.1794,
      "eval_samples_per_second": 335.438,
      "eval_steps_per_second": 5.278,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.025851409882307053,
      "learning_rate": 2.490417866302022e-05,
      "loss": 0.0141,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8973333333333333,
      "eval_loss": 0.40484535694122314,
      "eval_runtime": 10.9737,
      "eval_samples_per_second": 341.727,
      "eval_steps_per_second": 5.377,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.003626195015385747,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0045,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.3990466892719269,
      "eval_runtime": 11.8724,
      "eval_samples_per_second": 315.858,
      "eval_steps_per_second": 4.969,
      "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}