trainer_state.json

{
  "best_metric": 0.44133333333333336,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0445/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": 0.00028549422332891283,
      "loss": 3.3199,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22853333333333334,
      "eval_loss": 2.9531688690185547,
      "eval_runtime": 12.1677,
      "eval_samples_per_second": 308.192,
      "eval_steps_per_second": 4.849,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.72458267211914,
      "learning_rate": 0.00024415433369325305,
      "loss": 2.7936,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2717333333333333,
      "eval_loss": 2.717994451522827,
      "eval_runtime": 12.1611,
      "eval_samples_per_second": 308.36,
      "eval_steps_per_second": 4.852,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 17.574865341186523,
      "learning_rate": 0.0001841660235425207,
      "loss": 2.5438,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.32133333333333336,
      "eval_loss": 2.5300650596618652,
      "eval_runtime": 11.9741,
      "eval_samples_per_second": 313.177,
      "eval_steps_per_second": 4.927,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 28.175792694091797,
      "learning_rate": 0.00011741071341041153,
      "loss": 2.3026,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.36773333333333336,
      "eval_loss": 2.314760446548462,
      "eval_runtime": 12.1727,
      "eval_samples_per_second": 308.066,
      "eval_steps_per_second": 4.847,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.777849197387695,
      "learning_rate": 5.711011116874128e-05,
      "loss": 2.0219,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.40853333333333336,
      "eval_loss": 2.1806905269622803,
      "eval_runtime": 11.8469,
      "eval_samples_per_second": 316.539,
      "eval_steps_per_second": 4.98,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 21.73750114440918,
      "learning_rate": 1.5207490629826985e-05,
      "loss": 1.735,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4341333333333333,
      "eval_loss": 2.1214120388031006,
      "eval_runtime": 11.9102,
      "eval_samples_per_second": 314.857,
      "eval_steps_per_second": 4.954,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 36.14683151245117,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.53,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.44133333333333336,
      "eval_loss": 2.144158124923706,
      "eval_runtime": 12.1069,
      "eval_samples_per_second": 309.742,
      "eval_steps_per_second": 4.873,
      "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
}