trainer_state.json

{
  "best_metric": 0.39013333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0334/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": 8.887028694152832,
      "learning_rate": 0.00048519088401615274,
      "loss": 3.4149,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1632,
      "eval_loss": 3.1415278911590576,
      "eval_runtime": 5.7166,
      "eval_samples_per_second": 655.982,
      "eval_steps_per_second": 10.321,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.675899505615234,
      "learning_rate": 0.0004420155125428249,
      "loss": 2.937,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2,
      "eval_loss": 2.956073522567749,
      "eval_runtime": 4.3247,
      "eval_samples_per_second": 867.103,
      "eval_steps_per_second": 13.642,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 17.429214477539062,
      "learning_rate": 0.00037568023640968044,
      "loss": 2.7838,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.26666666666666666,
      "eval_loss": 2.721165895462036,
      "eval_runtime": 5.3207,
      "eval_samples_per_second": 704.79,
      "eval_steps_per_second": 11.089,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 23.968387603759766,
      "learning_rate": 0.00029418606892278544,
      "loss": 2.6061,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2906666666666667,
      "eval_loss": 2.6183831691741943,
      "eval_runtime": 4.4548,
      "eval_samples_per_second": 841.782,
      "eval_steps_per_second": 13.244,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 20.9128360748291,
      "learning_rate": 0.0002073624094068776,
      "loss": 2.4571,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3056,
      "eval_loss": 2.5570132732391357,
      "eval_runtime": 4.298,
      "eval_samples_per_second": 872.498,
      "eval_steps_per_second": 13.727,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 22.9094181060791,
      "learning_rate": 0.00012568147258031897,
      "loss": 2.2876,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.34613333333333335,
      "eval_loss": 2.3947432041168213,
      "eval_runtime": 4.5082,
      "eval_samples_per_second": 831.824,
      "eval_steps_per_second": 13.087,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 5.9164370142409884e-05,
      "loss": 2.1005,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.36746666666666666,
      "eval_loss": 2.29590106010437,
      "eval_runtime": 4.2973,
      "eval_samples_per_second": 872.631,
      "eval_steps_per_second": 13.729,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 19.19093132019043,
      "learning_rate": 1.5437430032376602e-05,
      "loss": 1.925,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.39013333333333333,
      "eval_loss": 2.260072708129883,
      "eval_runtime": 4.3814,
      "eval_samples_per_second": 855.899,
      "eval_steps_per_second": 13.466,
      "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
}