trainer_state.json

{
  "best_metric": 0.9269333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0308/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": 7.354092121124268,
      "learning_rate": 8.559608050575152e-05,
      "loss": 0.8006,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.48479726910591125,
      "eval_runtime": 7.1567,
      "eval_samples_per_second": 523.988,
      "eval_steps_per_second": 8.244,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.03682992607355118,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.2769,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.4955376088619232,
      "eval_runtime": 6.3315,
      "eval_samples_per_second": 592.278,
      "eval_steps_per_second": 9.319,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.039592444896698,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.1341,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8925333333333333,
      "eval_loss": 0.39644449949264526,
      "eval_runtime": 6.4684,
      "eval_samples_per_second": 579.744,
      "eval_steps_per_second": 9.121,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.012042325921356678,
      "learning_rate": 3.5104850050159014e-05,
      "loss": 0.0553,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8997333333333334,
      "eval_loss": 0.40008434653282166,
      "eval_runtime": 6.3575,
      "eval_samples_per_second": 589.852,
      "eval_steps_per_second": 9.28,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.01926332525908947,
      "learning_rate": 1.703789455194791e-05,
      "loss": 0.017,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9128,
      "eval_loss": 0.3415994644165039,
      "eval_runtime": 6.3642,
      "eval_samples_per_second": 589.235,
      "eval_steps_per_second": 9.271,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0049485135823488235,
      "learning_rate": 4.509176970441849e-06,
      "loss": 0.0021,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9216,
      "eval_loss": 0.3189493417739868,
      "eval_runtime": 6.4013,
      "eval_samples_per_second": 585.821,
      "eval_steps_per_second": 9.217,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0021229232661426067,
      "learning_rate": 1.6347729569543557e-10,
      "loss": 0.0006,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9269333333333334,
      "eval_loss": 0.3118165135383606,
      "eval_runtime": 7.4096,
      "eval_samples_per_second": 506.097,
      "eval_steps_per_second": 7.963,
      "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
}