trainer_state.json

{
  "best_metric": 0.9170666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0955/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": 2.574002574002574e-05,
      "loss": 0.8839,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.44116491079330444,
      "eval_runtime": 13.0979,
      "eval_samples_per_second": 286.306,
      "eval_steps_per_second": 4.505,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.146718146718147e-05,
      "loss": 0.2264,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.39625656604766846,
      "eval_runtime": 13.1325,
      "eval_samples_per_second": 285.551,
      "eval_steps_per_second": 4.493,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.116368293762207,
      "learning_rate": 1.718146718146718e-05,
      "loss": 0.0971,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.3449321687221527,
      "eval_runtime": 13.2749,
      "eval_samples_per_second": 282.488,
      "eval_steps_per_second": 4.444,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.02306542545557022,
      "learning_rate": 1.2895752895752896e-05,
      "loss": 0.0431,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9085333333333333,
      "eval_loss": 0.3359004557132721,
      "eval_runtime": 13.1232,
      "eval_samples_per_second": 285.754,
      "eval_steps_per_second": 4.496,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.963963508605957,
      "learning_rate": 8.61003861003861e-06,
      "loss": 0.0192,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9088,
      "eval_loss": 0.36387017369270325,
      "eval_runtime": 13.1497,
      "eval_samples_per_second": 285.177,
      "eval_steps_per_second": 4.487,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5365170240402222,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.0072,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9162666666666667,
      "eval_loss": 0.33439791202545166,
      "eval_runtime": 13.3445,
      "eval_samples_per_second": 281.014,
      "eval_steps_per_second": 4.421,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1043078601360321,
      "learning_rate": 3.861003861003861e-08,
      "loss": 0.0018,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9170666666666667,
      "eval_loss": 0.3384326696395874,
      "eval_runtime": 13.1719,
      "eval_samples_per_second": 284.698,
      "eval_steps_per_second": 4.479,
      "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
}