trainer_state.json

{
  "best_metric": 0.9413333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0567/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": 4.2900042900042906e-05,
      "loss": 0.7752,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.4160947799682617,
      "eval_runtime": 11.7474,
      "eval_samples_per_second": 319.22,
      "eval_steps_per_second": 5.022,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5072706937789917,
      "learning_rate": 3.575718575718576e-05,
      "loss": 0.2076,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8986666666666666,
      "eval_loss": 0.3278925120830536,
      "eval_runtime": 11.9181,
      "eval_samples_per_second": 314.646,
      "eval_steps_per_second": 4.95,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0385679267346859,
      "learning_rate": 2.8614328614328617e-05,
      "loss": 0.1001,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9045333333333333,
      "eval_loss": 0.33603954315185547,
      "eval_runtime": 11.5783,
      "eval_samples_per_second": 323.882,
      "eval_steps_per_second": 5.096,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0009225294343195856,
      "learning_rate": 2.147147147147147e-05,
      "loss": 0.0521,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9213333333333333,
      "eval_loss": 0.27580875158309937,
      "eval_runtime": 10.9856,
      "eval_samples_per_second": 341.357,
      "eval_steps_per_second": 5.371,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.021755507215857506,
      "learning_rate": 1.4328614328614328e-05,
      "loss": 0.0258,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9269333333333334,
      "eval_loss": 0.2707705497741699,
      "eval_runtime": 11.0607,
      "eval_samples_per_second": 339.039,
      "eval_steps_per_second": 5.334,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9874866008758545,
      "learning_rate": 7.1857571857571855e-06,
      "loss": 0.0099,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9368,
      "eval_loss": 0.2520907521247864,
      "eval_runtime": 11.6282,
      "eval_samples_per_second": 322.493,
      "eval_steps_per_second": 5.074,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.030373025685548782,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0037,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9413333333333334,
      "eval_loss": 0.25279712677001953,
      "eval_runtime": 11.6659,
      "eval_samples_per_second": 321.449,
      "eval_steps_per_second": 5.057,
      "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
}