Floating point exception (core dumped), but can't find core dump in /var/lib/systemd/coredump/

[junchen6072]

Where can I find the core dump?

[guillaumekln]

How did you get this error?

Can you share the steps to reproduce? (model size, transcription options, input file if possible)

[junchen6072]

How did you get this error?

Can you share the steps to reproduce? (model size, transcription options, input file if possible)

Hi, sorry I can't share the input file. This is annoying that prevents interactive debug with you awesome guys =.=

But the model is: model = WhisperModel('/home/whisper-large-v2-ct2/', device="cuda", compute_type="float16", cpu_threads=4, num_workers=2)

called with: segments, _ = model.transcribe(audio=audiof, word_timestamps=True)

[junchen6072]

I basically tried the notebook here: https://github.com/Majdoddin/nlp/blob/main/Pyannote_plays_and_Whisper_rhymes_v_2_0.ipynb

It's not happening on every of my input file though

[guillaumekln]

I found a case where this error happens and fixed it in this PR https://github.com/OpenNMT/CTranslate2/pull/1170.

I will update this issue when the fix is released and you will be able to test again.

[guillaumekln]

Can you update to ctranslate2>=3.12 and see if you still get the error?

[junchen6072]

thanks, I'll try today or tomorrow

[junchen6072]

sorry was busy with some deadlines. I wasn't able to run that library to repro today after tried for 10+mins, but since I'll not use it, I assume this is fixed. Thanks!