trainer_state.json

{
  "best_metric": 0.38506666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0817/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": 0.0004378753753753754,
      "loss": 3.43,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20533333333333334,
      "eval_loss": 3.0491912364959717,
      "eval_runtime": 12.0504,
      "eval_samples_per_second": 311.194,
      "eval_steps_per_second": 4.896,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 17.48873519897461,
      "learning_rate": 0.00037537537537537537,
      "loss": 2.9218,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.23893333333333333,
      "eval_loss": 2.8263051509857178,
      "eval_runtime": 11.9466,
      "eval_samples_per_second": 313.897,
      "eval_steps_per_second": 4.939,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.271120071411133,
      "learning_rate": 0.00031287537537537537,
      "loss": 2.7004,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2544,
      "eval_loss": 2.717020034790039,
      "eval_runtime": 12.1193,
      "eval_samples_per_second": 309.425,
      "eval_steps_per_second": 4.868,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 21.84552574157715,
      "learning_rate": 0.00025037537537537537,
      "loss": 2.5023,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2925333333333333,
      "eval_loss": 2.5717039108276367,
      "eval_runtime": 12.1108,
      "eval_samples_per_second": 309.641,
      "eval_steps_per_second": 4.872,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 29.305007934570312,
      "learning_rate": 0.0001878753753753754,
      "loss": 2.3033,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3234666666666667,
      "eval_loss": 2.515787124633789,
      "eval_runtime": 11.9764,
      "eval_samples_per_second": 313.117,
      "eval_steps_per_second": 4.926,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.286393165588379,
      "learning_rate": 0.0001253753753753754,
      "loss": 2.0893,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3589333333333333,
      "eval_loss": 2.344508171081543,
      "eval_runtime": 11.9648,
      "eval_samples_per_second": 313.419,
      "eval_steps_per_second": 4.931,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 27.87150764465332,
      "learning_rate": 6.287537537537538e-05,
      "loss": 1.8281,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3792,
      "eval_loss": 2.3091022968292236,
      "eval_runtime": 12.0153,
      "eval_samples_per_second": 312.102,
      "eval_steps_per_second": 4.91,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 5.630630630630631e-07,
      "loss": 1.4913,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.38506666666666667,
      "eval_loss": 2.4103713035583496,
      "eval_runtime": 12.0965,
      "eval_samples_per_second": 310.006,
      "eval_steps_per_second": 4.877,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}