trainer_state.json

{
  "best_metric": 0.9181333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0189/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": 0.3729734420776367,
      "learning_rate": 9.510675611750168e-05,
      "loss": 0.8594,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8053333333333333,
      "eval_loss": 0.6328060030937195,
      "eval_runtime": 5.0517,
      "eval_samples_per_second": 742.329,
      "eval_steps_per_second": 11.679,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.13322912197567e-05,
      "loss": 0.3611,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5304207801818848,
      "eval_runtime": 4.4849,
      "eval_samples_per_second": 836.143,
      "eval_steps_per_second": 13.155,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.08490660041570663,
      "learning_rate": 6.13230482918689e-05,
      "loss": 0.1955,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.44153058528900146,
      "eval_runtime": 4.4683,
      "eval_samples_per_second": 839.243,
      "eval_steps_per_second": 13.204,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.530058860778809,
      "learning_rate": 3.90711367817024e-05,
      "loss": 0.1014,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4436291456222534,
      "eval_runtime": 4.6264,
      "eval_samples_per_second": 810.557,
      "eval_steps_per_second": 12.753,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 20.38634490966797,
      "learning_rate": 1.8983820665631146e-05,
      "loss": 0.0436,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9058666666666667,
      "eval_loss": 0.3652040958404541,
      "eval_runtime": 4.5421,
      "eval_samples_per_second": 825.605,
      "eval_steps_per_second": 12.99,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.027014685794711113,
      "learning_rate": 5.0396392552081564e-06,
      "loss": 0.0129,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9117333333333333,
      "eval_loss": 0.34713470935821533,
      "eval_runtime": 5.659,
      "eval_samples_per_second": 662.66,
      "eval_steps_per_second": 10.426,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0013052340364083648,
      "learning_rate": 4.0869292992629095e-10,
      "loss": 0.0038,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9181333333333334,
      "eval_loss": 0.32724323868751526,
      "eval_runtime": 7.1795,
      "eval_samples_per_second": 522.317,
      "eval_steps_per_second": 8.218,
      "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
}