Instructions to use bofenghuang/whisper-small-cv11-french with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use bofenghuang/whisper-small-cv11-french with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="bofenghuang/whisper-small-cv11-french")# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("bofenghuang/whisper-small-cv11-french") model = AutoModelForMultimodalLM.from_pretrained("bofenghuang/whisper-small-cv11-french") - Notebooks
- Google Colab
- Kaggle
Not multilingual support
I tried fine tuning your model to add extra accents and new words but when I use it with RealTimeSTT for realtime transcription and set the preferred language to "fr" I get this warning message:
"RealTimeSTT: faster_whisper - WARNING - The current model is English-only but the language parameter is set to 'fr'; using 'en' instead."
Though it is not really an error so the transcription keeps going on without interrupting but out of that I figured that your might have fine tune your model based on whisper-small.en rather than whisper-small by mistake. But when using your model as main reference for real time transcription, it works well without any issue, it's just that if you try to finetune it with French to develop it further then the fine tuned model based on yours might not work.