Something wrong about image resize

[Genhajime]

Thank you for your great work!\ I tried to run "Quick Start", and got the following error. Would you please fix this bug?

Traceback (most recent call last):
  File "./pseudo-attributes/main.py", line 156, in <module>
    main()
  File "./pseudo-attributes/main.py", line 149, in main
    trainer.test_unbiased(epoch=epoch)
  File "./pseudo-attributes/trainer.py", line 234, in test_unbiased
    for data, target, biases, group, _, _, ids in test_bar:
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/tqdm/std.py", line 1195, in __iter__
    for obj in iterable:
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 652, in __next__
    data = self._next_data()
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 1347, in _next_data
    return self._process_data(data)
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 1373, in _process_data
    data.reraise()
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/_utils.py", line 461, in reraise
    raise exception
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/_utils/worker.py", line 302, in _worker_loop
    data = fetcher.fetch(index)
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
    return self.collate_fn(data)
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/_utils/collate.py", line 175, in default_collate
    return [default_collate(samples) for samples in transposed]  # Backwards compatibility.
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/_utils/collate.py", line 175, in <listcomp>
    return [default_collate(samples) for samples in transposed]  # Backwards compatibility.
  File "./miniconda3/envs/pseudo/lib/python3.10/site-packages/torch/utils/data/_utils/collate.py", line 140, in default_collate
    out = elem.new(storage).resize_(len(batch), *list(elem.size()))
RuntimeError: Trying to resize storage that is not resizable

[skynbe]

Hi, did you download celebA dataset at ${ROOT_PATH}/data/celebA? The directory should contain

./data/celebA
 └──img_align_celeba/
 └──list_attr_celeba.csv
 └──list_attr_celeba.txt
 └──list_eval_partition.csv
 └──list_eval_partition.txt

[Genhajime]

I appreciate your reply!

Actually, as downloading torchvision.datasets.CelebA is currently unavailable, I downloaded the data from the google drive directly. (https://drive.google.com/drive/folders/0B7EVK8r0v71pWEZsZE9oNnFzTm8?resourcekey=0-5BR16BdXnb8hVj6CNHKzLg) There are no celeba/ or *.csv in their folder, so I'm wondering how to get them.

[skynbe]

Please download the files from these two links:

https://www.kaggle.com/datasets/jessicali9530/celeba-dataset https://graal.ift.ulaval.ca/public/celeba/

(I found celeba/ folder is not required.)

[Genhajime]

All right, now the code works, thank you!

By the way, I have one more question. Could you tell me which epoch's model was used for the reported test accuracy on the paper?\ Just after whole training, or decided by validation accuracy?

[skynbe]

We selected the final model based on the validation accuracy.