Error in evaluateFromList() in SpeakerNet.py

[lobetb]

Similar to #150 , I get the following error :

Reading 4800 of 4914: 130.56 Hz, embedding size 512Traceback (most recent call last):
File "/home/ben/Desktop/voxceleb_trainer/./trainSpeakerNet.py", line 276, in
main()
File "/home/ben/Desktop/voxceleb_trainer/./trainSpeakerNet.py", line 272, in main
main_worker(0, None, args)
File "/home/ben/Desktop/voxceleb_trainer/./trainSpeakerNet.py", line 183, in main_worker
sc, lab, _ = trainer.evaluateFromList(**vars(args))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/SpeakerNet.py", line 170, in evaluateFromList
for idx, data in enumerate(test_loader):
^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/utils/data/dataloader.py", line 701, in next
data = self._next_data()
^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/utils/data/dataloader.py", line 1465, in _next_data
return self._process_data(data)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/utils/data/dataloader.py", line 1491, in _process_data
data.reraise()
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/_utils.py", line 715, in reraise
raise exception
RuntimeError: Caught RuntimeError in DataLoader worker process 4.
Original Traceback (most recent call last):
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/utils/data/_utils/worker.py", line 351, in _worker_loop
data = fetcher.fetch(index) # type: ignore[possibly-undefined]
^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
~~~~~~~~~~~~^^^^^
File "/home/ben/Desktop/voxceleb_trainer/DatasetLoader.py", line 179, in getitem
audio = loadWAV(os.path.join(self.test_path,self.test_list[index]), self.max_frames, evalmode=True, num_eval=self.num_eval)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/DatasetLoader.py", line 32, in loadWAV
audio, sample_rate = soundfile.read(filename)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/soundfile.py", line 256, in read
with SoundFile(file, 'r', samplerate, channels,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/soundfile.py", line 629, in init
self._file = self._open(file, mode_int, closefd)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/soundfile.py", line 1183, in _open
_error_check(_snd.sf_error(file_ptr),
File "/home/ben/Desktop/voxceleb_trainer/env/lib/python3.12/site-packages/soundfile.py", line 1357, in _error_check
raise RuntimeError(prefix + _ffi.string(err_str).decode('utf-8', 'replace'))
RuntimeError: Error opening 'data/voxceleb1/id10270': System error.

While using the command :
python ./trainSpeakerNet.py --eval --model ResNetSE34L --log_input True --trainfunc angleproto --save_path data/voxceleb1 --eval_frames 400 --initial_model baseline_lite_ap.model

I noticed that it says that it has 4914 lines to read while there is only 4874 files (and lines in test_list.txt).
I have tried to put all the wav files in the "voxceleb1" base folder instead of distributed in subfolders as the dataset is originally to no avail.

using

• Python 3.12.7
• torch 2.5.0+rocm6.2