trainer_state.json

{
  "best_metric": 0.9176,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0988/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": 4.755337805875084e-05,
      "loss": 0.788,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8490666666666666,
      "eval_loss": 0.45805782079696655,
      "eval_runtime": 13.0642,
      "eval_samples_per_second": 287.045,
      "eval_steps_per_second": 4.516,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 150.4292755126953,
      "learning_rate": 4.063987381465584e-05,
      "loss": 0.2491,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.44456470012664795,
      "eval_runtime": 13.9892,
      "eval_samples_per_second": 268.065,
      "eval_steps_per_second": 4.218,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 13.266216278076172,
      "learning_rate": 3.0628700751103485e-05,
      "loss": 0.1258,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8845333333333333,
      "eval_loss": 0.37671568989753723,
      "eval_runtime": 12.9912,
      "eval_samples_per_second": 288.656,
      "eval_steps_per_second": 4.542,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.00362362340092659,
      "learning_rate": 1.950269447231056e-05,
      "loss": 0.0586,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.3972715437412262,
      "eval_runtime": 12.9836,
      "eval_samples_per_second": 288.827,
      "eval_steps_per_second": 4.544,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0023648154456168413,
      "learning_rate": 9.465496973304394e-06,
      "loss": 0.0234,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.3580321669578552,
      "eval_runtime": 12.9793,
      "eval_samples_per_second": 288.922,
      "eval_steps_per_second": 4.546,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.03040633350610733,
      "learning_rate": 2.5050983169121382e-06,
      "loss": 0.0059,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9152,
      "eval_loss": 0.3451319634914398,
      "eval_runtime": 12.9656,
      "eval_samples_per_second": 289.226,
      "eval_steps_per_second": 4.55,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.24721118807792664,
      "learning_rate": 9.082071983079754e-11,
      "loss": 0.0022,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9176,
      "eval_loss": 0.3415372371673584,
      "eval_runtime": 13.1221,
      "eval_samples_per_second": 285.778,
      "eval_steps_per_second": 4.496,
      "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
}