trainer_state.json

{
  "best_metric": 0.904,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0789/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": 8.559608050575152e-05,
      "loss": 0.8806,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8184,
      "eval_loss": 0.5824146866798401,
      "eval_runtime": 12.0204,
      "eval_samples_per_second": 311.97,
      "eval_steps_per_second": 4.908,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04923570528626442,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.288,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8461333333333333,
      "eval_loss": 0.5203445553779602,
      "eval_runtime": 11.9934,
      "eval_samples_per_second": 312.672,
      "eval_steps_per_second": 4.919,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3931562900543213,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.1385,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.5705666542053223,
      "eval_runtime": 11.9553,
      "eval_samples_per_second": 313.668,
      "eval_steps_per_second": 4.935,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 3.516402310353216e-05,
      "loss": 0.0659,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.5238242149353027,
      "eval_runtime": 11.876,
      "eval_samples_per_second": 315.762,
      "eval_steps_per_second": 4.968,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.16663549840450287,
      "learning_rate": 1.708543859906803e-05,
      "loss": 0.0138,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.888,
      "eval_loss": 0.5051289796829224,
      "eval_runtime": 12.0173,
      "eval_samples_per_second": 312.051,
      "eval_steps_per_second": 4.91,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.012484543025493622,
      "learning_rate": 4.53567532968734e-06,
      "loss": 0.0017,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9037333333333334,
      "eval_loss": 0.45181405544281006,
      "eval_runtime": 11.9368,
      "eval_samples_per_second": 314.154,
      "eval_steps_per_second": 4.943,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.032232195138931274,
      "learning_rate": 3.6782363693366185e-10,
      "loss": 0.0002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.4515075981616974,
      "eval_runtime": 12.2679,
      "eval_samples_per_second": 305.675,
      "eval_steps_per_second": 4.809,
      "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
}