Open Dannypeja opened 9 months ago
Weird. I am only having issues when I use it inside Docker Container. Natively it is fine. Any Ideas?
Did you accept terms for the new 3.0 model cards?
If 3.0 isn't working for you (it is slower than 2.1 as discussed #499), then try to go back to 2.1
model = whisperx.DiarizationPipeline(model_name='pyannote/speaker-diarization@2.1', use_auth_token=YOUR_AUTH_TOKEN, device='cuda')
I did not. Now I did. Still same error. Interestingly it works on host, worked even before I accepted terms. But inside a docker container it does not work.
How can I go back to 2.1 when using whisperx cli?
Okay not only do I need to accept 3.0 diarization but also segmentation: https://hf.co/pyannote/segmentation-3.0
Wow, a full day just for this.
Yes it seems to be slower. I would still be interested to learn how I can revert to 2.1 using whisperx CLI
Yes, both segmentation and diarization. You can try my hack in #499 to make 3.0 work fast.
I'll try it, tanks! Is 3.0 faster that 2.1? Otherwise I think I'll stay at 2.1 cause it works fine..
Slightly faster from the limited samples I saw. But I am sticking with 2.1 until no hack is needed for 3.0.
Currently readme.md appears to link to the previous segmentation model on Huggingface:
https://huggingface.co/pyannote/segmentation
instead of:
I believe you have to accept both of these two conditions - https://huggingface.co/pyannote/segmentation-3.0 and https://huggingface.co/pyannote/speaker-diarization-3.0
I believe you have to accept both of these two conditions - https://huggingface.co/pyannote/segmentation-3.0 and https://huggingface.co/pyannote/speaker-diarization-3.0
hi, I accepted both , but I am having same problem. Is there anything that I can do more?
A newer diarization model (3.1) is being used by whisperx. Accept this one instead https://huggingface.co/pyannote/speaker-diarization-3.1
I currently accept segmentation, segmentation3.0, diarisation3.0, diarizaton3.1 and still have no permissions
Are we having this again? Please confirm that I am not stupid. I have been desperate for 6 hours. I thought it was my fault somehow.
490
pyannote related issue
I am getting this