Open TokerX opened 1 year ago
The torchaudio one is intended on windows (since soundfile is the default one on windows). So this shouldn't be the issue. The raised ValueError is a bit misleading - it is indeed a permission error. Just to double check that it is not a permission error (admin should fix it, but I'm not too familiar with windows so I'm not certain): can you try to manually set the permissions of C:\Users\svenc\.cache\huggingface
?
Since it's failing on symlink creation, the other directory ..\..\blobs\bb3285bc209d674e3f88646bdfd327bfe43b60da
could be the issue too. This is also located in the same Cache-Dir, so setting C:\Users\svenc\.cache\huggingface
should be enough.
I get the following errors upon running whisperX
C:\Windows\System32>whisperx --model large-v2 --language nl "F:\Movies\Ad Fundum (1993)\Ad Fundum (1993).avi" --compute_type float32 The torchaudio backend is switched to 'soundfile'. Note that 'sox_io' is not supported on Windows. The torchaudio backend is switched to 'soundfile'. Note that 'sox_io' is not supported on Windows. Lightning automatically upgraded your loaded checkpoint from v1.5.4 to v2.0.2. To apply the upgrade to your files permanently, run
python -m pytorch_lightning.utilities.upgrade_checkpoint --file C:\Users\svenc.cache\torch\whisperx-vad-segmentation.bin` Model was trained with pyannote.audio 0.0.1, yours is 2.1.1. Bad things might happen unless you revert pyannote.audio to 0.x. Model was trained with torch 1.10.0+cu102, yours is 2.0.0+cpu. Bad things might happen unless you revert torch to 1.x.During handling of the above exception, another exception occurred:
Traceback (most recent call last): File "", line 198, in _run_module_as_main
File "", line 88, in _run_code
File "C:\Users\svenc\AppData\Local\Programs\Python\Python311\Scripts\whisperx.exe__main__.py", line 7, in
File "C:\Users\svenc\AppData\Local\Programs\Python\Python311\Lib\site-packages\whisperx\transcribe.py", line 166, in cli
align_model, align_metadata = load_align_model(align_language, device, model_name=align_model)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\svenc\AppData\Local\Programs\Python\Python311\Lib\site-packages\whisperx\alignment.py", line 73, in load_align_model
raise ValueError(f'The chosen align_model "{model_name}" could not be found in huggingface (https://huggingface.co/models) or torchaudio (https://pytorch.org/audio/stable/pipelines.html#id14)')
ValueError: The chosen align_model "jonatasgrosman/wav2vec2-large-xlsr-53-dutch" could not be found in huggingface (https://huggingface.co/models) or torchaudio (https://pytorch.org/audio/stable/pipelines.html#id14)`
The last one talks about torchaudio as well, but seeing as everything else is about huggingface I guess that's where the problem is.
"PermissionError: [WinError 5] Access is denied: " makes it seem like a Windows thing or something, but I'm running as admin so it's not a question of having rights.