trainer_state.json

{
  "best_metric": 0.8912,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0053/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": 2.85232922088503e-05,
      "loss": 2.3014,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7733333333333333,
      "eval_loss": 1.08161199092865,
      "eval_runtime": 12.6519,
      "eval_samples_per_second": 296.398,
      "eval_steps_per_second": 4.663,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.577245712280273,
      "learning_rate": 2.4368144166557088e-05,
      "loss": 0.7317,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.6208202838897705,
      "eval_runtime": 12.5369,
      "eval_samples_per_second": 299.117,
      "eval_steps_per_second": 4.706,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.390222072601318,
      "learning_rate": 1.8357520279328874e-05,
      "loss": 0.3988,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.4942691922187805,
      "eval_runtime": 12.7578,
      "eval_samples_per_second": 293.938,
      "eval_steps_per_second": 4.625,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.297302007675171,
      "learning_rate": 1.1681898323495617e-05,
      "loss": 0.2433,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8797333333333334,
      "eval_loss": 0.4492838978767395,
      "eval_runtime": 13.7089,
      "eval_samples_per_second": 273.544,
      "eval_steps_per_second": 4.304,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8554706573486328,
      "learning_rate": 5.6634670985405755e-06,
      "loss": 0.1505,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.41320839524269104,
      "eval_runtime": 12.9038,
      "eval_samples_per_second": 290.612,
      "eval_steps_per_second": 4.572,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.8068761825561523,
      "learning_rate": 1.4942507197812133e-06,
      "loss": 0.1006,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.4099847376346588,
      "eval_runtime": 12.5127,
      "eval_samples_per_second": 299.696,
      "eval_steps_per_second": 4.715,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 7.834527015686035,
      "learning_rate": 1.362311416175377e-11,
      "loss": 0.0795,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.41191965341567993,
      "eval_runtime": 12.6466,
      "eval_samples_per_second": 296.522,
      "eval_steps_per_second": 4.665,
      "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
}