trainer_state.json

{
  "best_metric": 0.9490666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0712/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": 3.071072578430176,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.6127,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.2619973123073578,
      "eval_runtime": 11.0933,
      "eval_samples_per_second": 338.043,
      "eval_steps_per_second": 5.319,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.074691295623779,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.1281,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9290666666666667,
      "eval_loss": 0.24974454939365387,
      "eval_runtime": 11.1632,
      "eval_samples_per_second": 335.926,
      "eval_steps_per_second": 5.285,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.44197511672973633,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.0622,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9330666666666667,
      "eval_loss": 0.2485334277153015,
      "eval_runtime": 11.8318,
      "eval_samples_per_second": 316.942,
      "eval_steps_per_second": 4.987,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06310944259166718,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0286,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9376,
      "eval_loss": 0.2503167986869812,
      "eval_runtime": 10.9793,
      "eval_samples_per_second": 341.551,
      "eval_steps_per_second": 5.374,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.7620584964752197,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0126,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9448,
      "eval_loss": 0.2456582635641098,
      "eval_runtime": 10.7266,
      "eval_samples_per_second": 349.599,
      "eval_steps_per_second": 5.5,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0487663559615612,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.0054,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9488,
      "eval_loss": 0.24018853902816772,
      "eval_runtime": 11.4669,
      "eval_samples_per_second": 327.027,
      "eval_steps_per_second": 5.145,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0027,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9490666666666666,
      "eval_loss": 0.2316489964723587,
      "eval_runtime": 11.2573,
      "eval_samples_per_second": 333.117,
      "eval_steps_per_second": 5.241,
      "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
}