trainer_state.json

{
  "best_metric": 0.3738666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0025/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 0.00048527961022201525,
      "loss": 3.4882,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.18613333333333335,
      "eval_loss": 3.0864968299865723,
      "eval_runtime": 5.6176,
      "eval_samples_per_second": 667.545,
      "eval_steps_per_second": 10.503,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.44053840637207,
      "learning_rate": 0.0004421832247293072,
      "loss": 3.0295,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.21226666666666666,
      "eval_loss": 2.983042001724243,
      "eval_runtime": 5.3147,
      "eval_samples_per_second": 705.585,
      "eval_steps_per_second": 11.101,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.490388870239258,
      "learning_rate": 0.00037590670601192445,
      "loss": 2.851,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.24586666666666668,
      "eval_loss": 2.853865146636963,
      "eval_runtime": 5.2233,
      "eval_samples_per_second": 717.937,
      "eval_steps_per_second": 11.296,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.747262954711914,
      "learning_rate": 0.00029444398036442514,
      "loss": 2.6833,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2754666666666667,
      "eval_loss": 2.6955316066741943,
      "eval_runtime": 5.6865,
      "eval_samples_per_second": 659.452,
      "eval_steps_per_second": 10.375,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 21.383445739746094,
      "learning_rate": 0.00020762065476168375,
      "loss": 2.502,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3194666666666667,
      "eval_loss": 2.572685480117798,
      "eval_runtime": 5.6432,
      "eval_samples_per_second": 664.52,
      "eval_steps_per_second": 10.455,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 23.32973861694336,
      "learning_rate": 0.00012590890364720656,
      "loss": 2.3042,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.356,
      "eval_loss": 2.4144418239593506,
      "eval_runtime": 6.2084,
      "eval_samples_per_second": 604.021,
      "eval_steps_per_second": 9.503,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 15.162907600402832,
      "learning_rate": 5.9164370142409884e-05,
      "loss": 2.0625,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3682666666666667,
      "eval_loss": 2.3916568756103516,
      "eval_runtime": 5.3887,
      "eval_samples_per_second": 695.9,
      "eval_steps_per_second": 10.949,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 1.5528220942544385e-05,
      "loss": 1.793,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.3738666666666667,
      "eval_loss": 2.432835578918457,
      "eval_runtime": 5.5729,
      "eval_samples_per_second": 672.903,
      "eval_steps_per_second": 10.587,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}