metadata
title: Speech To Speech Translation
emoji: 🏆
colorFrom: pink
colorTo: indigo
sdk: gradio
sdk_version: 3.36.1
app_file: app.py
pinned: false
Part of the HuggingFace Audio Processing course.
This is a Gradio wrapper around a (X -> Malay) speech2speech pipeline, where X is any language supported by
openai/whisper-base.
The TTS model used is facebook/mms-tts-zlm, a pretrained checkpoint for speech in Malay which is part of their Massively Multilingual Speech project. The underlying architecture is based on VITS, which generates waveforms directly and does not need a seperate vocoder.
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference