CUDA error: the launch timed out and was terminated

[federicodalzotto]

Hi all.
Let me preface this by saying that I don't understand anything about this, but I can do all the tests you want if you explain clearly what to do.
I have installed “Standalone Faster-Whisper-XXL r245.2” on a GPU virtual machine with NVIDIA A40-1Q version 538.95
I launched this command: C:\Whisper\faster-whisper-xxl.exe "C:\Whisper\Input\MySpeech.mp3" --language English --model large-v2 --output_dir "C:\Whisper\Output" --output_format txt --diarize pyannote_v3.1 --num_speakers=4

And I had this output error:

Traceback (most recent call last):
File "D:\whisper-fast-XXL\__main__.py", line 2245, in <module>
File "D:\whisper-fast-XXL\__main__.py", line 1925, in cli
File "D:\whisper-fast-XXL\__main__.py", line 405, in get_diarization
File "faster_whisper\diarization.py", line 90, in run_model
File "pyannote\audio\core\pipeline.py", line 326, in __call__
File "pyannote\audio\pipelines\speaker_diarization.py", line 524, in apply
File "pyannote\audio\pipelines\speaker_diarization.py", line 348, in get_embeddings
File "pyannote\audio\pipelines\speaker_verification.py", line 707, in __call__
RuntimeError: CUDA error: the launch timed out and was terminated
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
[PYI-16188:ERROR] Failed to execute script '__main__' due to unhandled exception!

This is the second Virtual Machine with same GPU configuration where I find the very same issue.

Can you please help me in having this thing working?
Federico

[Purfview]

Moved to the issues: #411