trainer_state.json

{
  "best_metric": 0.8784,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0290/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 5e-05,
      "loss": 0.7805,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.46053552627563477,
      "eval_runtime": 6.177,
      "eval_samples_per_second": 607.093,
      "eval_steps_per_second": 9.552,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.9801025390625,
      "learning_rate": 5e-05,
      "loss": 0.2149,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.43906307220458984,
      "eval_runtime": 6.1637,
      "eval_samples_per_second": 608.402,
      "eval_steps_per_second": 9.572,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.1171,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.4917280375957489,
      "eval_runtime": 6.2115,
      "eval_samples_per_second": 603.716,
      "eval_steps_per_second": 9.498,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.07685737311840057,
      "learning_rate": 5e-05,
      "loss": 0.0773,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.5179932713508606,
      "eval_runtime": 6.624,
      "eval_samples_per_second": 566.123,
      "eval_steps_per_second": 8.907,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.00021168863167986274,
      "learning_rate": 5e-05,
      "loss": 0.0633,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.47392964363098145,
      "eval_runtime": 6.2123,
      "eval_samples_per_second": 603.637,
      "eval_steps_per_second": 9.497,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.09366428107023239,
      "learning_rate": 5e-05,
      "loss": 0.0553,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8784,
      "eval_loss": 0.513299822807312,
      "eval_runtime": 6.2267,
      "eval_samples_per_second": 602.247,
      "eval_steps_per_second": 9.475,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}