alexk2037

@nshmyrev If you have any ideas for how to improve the handling of partial results I am more than happy to test them out.

Hello, I was trying to apply profanity filtering (using the [profanity-filter python library](https://github.com/rominf/profanity-filter)) and I have some confusion about how Vosk produces partial and full transcripts. In the image below,...

As for a list of swear words, there is an audio dataset called TAPAD that contains audio files of numerous swear words: https://github.com/theabuseproject/tapad The files are in `.mp3` format, but...

Update on the profanity filter -------------------------------- I decided to use two libraries, [profanity-filter](https://pypi.org/project/profanity-filter/) and [alt-profanity-check](https://pypi.org/project/alt-profanity-check/) because the combination performs faster than using profanity filter alone. I tested two conditions on...

## Profanity Filter Implementation Update `vosk_server/websocket/asr_server.py` ```python #!/usr/bin/env python3 import json import os import sys import asyncio import pathlib import websockets import concurrent.futures import logging from vosk import Model, SpkModel,...