Files
pyannote-audio/FAQ.md
T
2022-11-29 18:14:49 +01:00

2.7 KiB

Frequently asked questions

How does one capitalize and pronounce the name of this awesome library?

📝 Written in lower case: pyannote.audio (or pyannote if you are lazy). Not PyAnnote nor PyAnnotate (sic). 📢 Pronounced like the french verb pianoter. pi like in piano, not py like in python. 🎹 pianoter means to play the piano (hence the logo 🤯).

Can I use gated models (and pipelines) offline?

Short answer: yes, see this tutorial for models and that one for pipelines.

Long answer: gating models and pipelines allows me to know a bit more about pyannote.audio user base and eventually help me write grant proposals to make pyannote.audio even better. So, please fill gating forms as precisely as possible.

For instance, before gating pyannote/speaker-diarization, I had no idea that so many people were relying on it in production. Hint: sponsors are more than welcome! Maintaining open source libraries is time consuming.

That being said, this whole authentication process does not prevent you from using official pyannote.audio models offline (i.e. without going through the authentication process in every docker run ... or whatever you are using in production): see this tutorial for models and that one for pipelines.

Pretrained pipelines do not produce good results on my data. What can I do?

  1. Annotate dozens of conversations manually and separate them into development and test subsets in pyannote.database.
  2. Optimize the hyper-parameters of the pretained pipeline using the development set. If performance is still not good enough, go to step 3.
  3. Annotate hundreds of conversations manually and set them up as training subset in pyannote.database.
  4. Fine-tune the models (on which the pipeline relies) using the training set.
  5. Optimize the hyper-parameters of the pipeline using the fine-tuned models using the development set. If performance is still not good enough, go back to step 3.