trainer_state.json

{
  "best_metric": 0.41146666666666665,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0045/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 0.0004380630630630631,
      "loss": 3.5253,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.18746666666666667,
      "eval_loss": 3.066896915435791,
      "eval_runtime": 6.49,
      "eval_samples_per_second": 577.811,
      "eval_steps_per_second": 9.091,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.248388290405273,
      "learning_rate": 0.0003755630630630631,
      "loss": 2.9821,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.23946666666666666,
      "eval_loss": 2.8167800903320312,
      "eval_runtime": 6.2728,
      "eval_samples_per_second": 597.821,
      "eval_steps_per_second": 9.406,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 22.379396438598633,
      "learning_rate": 0.0003130630630630631,
      "loss": 2.7504,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2770666666666667,
      "eval_loss": 2.6739094257354736,
      "eval_runtime": 6.363,
      "eval_samples_per_second": 589.347,
      "eval_steps_per_second": 9.272,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 21.272457122802734,
      "learning_rate": 0.0002505630630630631,
      "loss": 2.5736,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.31733333333333336,
      "eval_loss": 2.4984865188598633,
      "eval_runtime": 6.4913,
      "eval_samples_per_second": 577.696,
      "eval_steps_per_second": 9.089,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 36.86085891723633,
      "learning_rate": 0.00018806306306306305,
      "loss": 2.3706,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3392,
      "eval_loss": 2.3726093769073486,
      "eval_runtime": 6.6156,
      "eval_samples_per_second": 566.844,
      "eval_steps_per_second": 8.918,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 26.497859954833984,
      "learning_rate": 0.00012556306306306305,
      "loss": 2.164,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3834666666666667,
      "eval_loss": 2.2380142211914062,
      "eval_runtime": 6.3675,
      "eval_samples_per_second": 588.926,
      "eval_steps_per_second": 9.266,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 31.70774269104004,
      "learning_rate": 6.306306306306306e-05,
      "loss": 1.9415,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4026666666666667,
      "eval_loss": 2.1743690967559814,
      "eval_runtime": 6.4665,
      "eval_samples_per_second": 579.913,
      "eval_steps_per_second": 9.124,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 7.507507507507508e-07,
      "loss": 1.6791,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.41146666666666665,
      "eval_loss": 2.143007755279541,
      "eval_runtime": 6.9574,
      "eval_samples_per_second": 538.995,
      "eval_steps_per_second": 8.48,
      "step": 2664
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}