DhivehiSTT

This repository contains code to run STT model used in the DV-Subs project by Dhivehi.ai

Setup

1. git clone https://github.com/ashraq1455/DhivehiSTT.git
2. pip install -r requirements.txt
3. Download the model from here
4. Extract and copy wav2vec_traced_quantized.pt and vocab.json to models directory

How to use

from stt import transcribe
transcription = transcribe(audio_bytes)

If you want to use with http, run python main.py and send a POST request with the audio to predict/ endpoint.