trainer_state.json

{
  "best_metric": 0.9373333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0577/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": 1.1005148887634277,
      "learning_rate": 6.733578363789503e-05,
      "loss": 0.7307,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.30461689829826355,
      "eval_runtime": 11.0295,
      "eval_samples_per_second": 339.998,
      "eval_steps_per_second": 5.349,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.512343168258667,
      "learning_rate": 5.974873734152915e-05,
      "loss": 0.1421,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9250666666666667,
      "eval_loss": 0.23986020684242249,
      "eval_runtime": 11.1484,
      "eval_samples_per_second": 336.37,
      "eval_steps_per_second": 5.292,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.045750878751277924,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.0562,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9266666666666666,
      "eval_loss": 0.24691694974899292,
      "eval_runtime": 10.9506,
      "eval_samples_per_second": 342.446,
      "eval_steps_per_second": 5.388,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.02840917557477951,
      "learning_rate": 3.5e-05,
      "loss": 0.0204,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9341333333333334,
      "eval_loss": 0.24977177381515503,
      "eval_runtime": 10.9382,
      "eval_samples_per_second": 342.837,
      "eval_steps_per_second": 5.394,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.038862645626068115,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.0059,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.23659314215183258,
      "eval_runtime": 10.9656,
      "eval_samples_per_second": 341.978,
      "eval_steps_per_second": 5.38,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02829059585928917,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.0021,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.2557178735733032,
      "eval_runtime": 11.0369,
      "eval_samples_per_second": 339.769,
      "eval_steps_per_second": 5.346,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.6800339785621635e-06,
      "loss": 0.0015,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9373333333333334,
      "eval_loss": 0.24724674224853516,
      "eval_runtime": 10.9138,
      "eval_samples_per_second": 343.601,
      "eval_steps_per_second": 5.406,
      "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}