trainer_state.json

{
  "best_metric": 0.9101333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0774/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": 1.0921344757080078,
      "learning_rate": 9.507764069616767e-05,
      "loss": 0.9372,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8197333333333333,
      "eval_loss": 0.6061252355575562,
      "eval_runtime": 11.176,
      "eval_samples_per_second": 335.541,
      "eval_steps_per_second": 5.279,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.017167868092656136,
      "learning_rate": 8.122714722185695e-05,
      "loss": 0.3359,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.5409581661224365,
      "eval_runtime": 11.0545,
      "eval_samples_per_second": 339.227,
      "eval_steps_per_second": 5.337,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 42.4278450012207,
      "learning_rate": 6.119173426442958e-05,
      "loss": 0.1569,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.5154885053634644,
      "eval_runtime": 10.8977,
      "eval_samples_per_second": 344.109,
      "eval_steps_per_second": 5.414,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 3.900538894462112e-05,
      "loss": 0.079,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8773333333333333,
      "eval_loss": 0.45871198177337646,
      "eval_runtime": 10.8705,
      "eval_samples_per_second": 344.972,
      "eval_steps_per_second": 5.428,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.004699366167187691,
      "learning_rate": 1.893099394660879e-05,
      "loss": 0.0244,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9010666666666667,
      "eval_loss": 0.4241147041320801,
      "eval_runtime": 10.8604,
      "eval_samples_per_second": 345.293,
      "eval_steps_per_second": 5.433,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0029818187467753887,
      "learning_rate": 5.0101966338242764e-06,
      "loss": 0.0045,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9072,
      "eval_loss": 0.3732217252254486,
      "eval_runtime": 10.8882,
      "eval_samples_per_second": 344.409,
      "eval_steps_per_second": 5.419,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.008799640461802483,
      "learning_rate": 1.8164143966159507e-10,
      "loss": 0.0013,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9101333333333333,
      "eval_loss": 0.3649348318576813,
      "eval_runtime": 10.9406,
      "eval_samples_per_second": 342.76,
      "eval_steps_per_second": 5.393,
      "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
}