trainer_state.json

{
  "best_metric": 0.9269333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0974/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": 0.2694445252418518,
      "learning_rate": 2.572715572715573e-05,
      "loss": 0.8009,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8898666666666667,
      "eval_loss": 0.348587304353714,
      "eval_runtime": 8.2957,
      "eval_samples_per_second": 452.039,
      "eval_steps_per_second": 7.112,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.025404930114746,
      "learning_rate": 2.1441441441441442e-05,
      "loss": 0.1982,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.28767305612564087,
      "eval_runtime": 6.8858,
      "eval_samples_per_second": 544.596,
      "eval_steps_per_second": 8.568,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.551675319671631,
      "learning_rate": 1.715572715572716e-05,
      "loss": 0.0816,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9122666666666667,
      "eval_loss": 0.2928467392921448,
      "eval_runtime": 7.1647,
      "eval_samples_per_second": 523.397,
      "eval_steps_per_second": 8.235,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 23.90164566040039,
      "learning_rate": 1.287001287001287e-05,
      "loss": 0.0377,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9176,
      "eval_loss": 0.29722458124160767,
      "eval_runtime": 7.0525,
      "eval_samples_per_second": 531.729,
      "eval_steps_per_second": 8.366,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1511356681585312,
      "learning_rate": 8.584298584298585e-06,
      "loss": 0.0137,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9218666666666666,
      "eval_loss": 0.2734692692756653,
      "eval_runtime": 6.7688,
      "eval_samples_per_second": 554.014,
      "eval_steps_per_second": 8.716,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0006047762581147254,
      "learning_rate": 4.298584298584299e-06,
      "loss": 0.0037,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.92,
      "eval_loss": 0.2928309440612793,
      "eval_runtime": 6.7715,
      "eval_samples_per_second": 553.793,
      "eval_steps_per_second": 8.713,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0771004781126976,
      "learning_rate": 1.2870012870012871e-08,
      "loss": 0.0014,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9269333333333334,
      "eval_loss": 0.2835957109928131,
      "eval_runtime": 6.9805,
      "eval_samples_per_second": 537.21,
      "eval_steps_per_second": 8.452,
      "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
}