Error running evaluation for google_whale

[alpha-tango]

I got an error running evaluation for out-of-the-box google whale on @vskode 's data:

running inference on test data:   2%|▏         | 28/1600 [00:26<19:57,  1.31it/s]/gpfs2/scratch/atpope/bioacoustics/acodet/acodet/torch_data.py:66: UserWarning: PySoundFile failed. Trying audioread instead.
  wave, sr = lb.load(
/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/librosa/core/audio.py:184: FutureWarning: librosa.core.audio.__audioread_load
	Deprecated as of librosa version 0.10.0.
	It will be removed in librosa version 1.0.
  y, sr_native = __audioread_load(path, offset, duration, dtype)
running inference on test data:   5%|▍         | 79/1600 [01:05<20:58,  1.21it/s]
Traceback (most recent call last):
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/run.py", line 85, in <module>
    main()
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/run.py", line 74, in main
    evaluate(train_date=conf.LOAD_CKPT_PATH, **kwargs)
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/acodet/evaluate.py", line 67, in evaluate
    for idx, tuple in tqdm(
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/tqdm/std.py", line 1181, in __iter__
    for obj in iterable:
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 741, in __next__
    data = self._next_data()
           ^^^^^^^^^^^^^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1548, in _next_data
    return self._process_data(data, worker_id)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 1586, in _process_data
    data.reraise()
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/_utils.py", line 775, in reraise
    raise exception
ValueError: Caught ValueError in DataLoader worker process 15.
Original Traceback (most recent call last):
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/_utils/worker.py", line 358, in _worker_loop
    data = fetcher.fetch(index)  # type: ignore[possibly-undefined]
           ^^^^^^^^^^^^^^^^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 54, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 54, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
            ~~~~~~~~~~~~^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/acodet/torch_data.py", line 90, in __getitem__
    lb.util.fix_length(
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/librosa/util/utils.py", line 575, in fix_length
    return np.pad(data, lengths, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/gpfs2/scratch/atpope/bioacoustics/acodet/.venv/lib/python3.11/site-packages/numpy/lib/arraypad.py", line 819, in pad
    raise ValueError(
ValueError: can't extend empty axis 0 using modes other than 'constant' or 'empty'

[alpha-tango]

I got this error also for the US_Bermuda location subset of the data; all other subsets (Iceland, Scotland, Caribbean) ran evaluation successfully.

[vskode]

I think this could be coming from the fact that there are no negatives in some locations. I know that this was the case for scotland but I also don't think I ever evaluated it alone. So you could either add some negatives manually or just avoid evaluating these areas alone.

[ntraft]

Actually, this is an error in loading an audio file using the librosa library. I have not dug in deeper to figure out which wav files it's tripping over.