Instructions to use ARTPARK-IISc/Vaani-FastConformer-Multilingual with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use ARTPARK-IISc/Vaani-FastConformer-Multilingual with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("ARTPARK-IISc/Vaani-FastConformer-Multilingual") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Punctuation Support
#2
by sumit9821 - opened
Any guide on how to handle punctuation of the transcript in realtime, it seems punctuation is not availble in these(ASR) models.
I am running multilingual model
I think you could either fine-tune the model (after adding that to the vocabulary), or use a post-processing model/logic to include punctuation.