A relatively basic setup for transcribing Livestreams on Twitch using OpenAI's Whisper and Streamlink.
- Install ffmpeg with
sudo apt install ffmpeg
,choco install ffmpeg
... - Install Streamlink with
pip install streamlink
- Create a virtual environment
pip install -r requirements.txt
python main.py <twitch channel>
Twitch doesn't have a built-in transcription feature, and I wanted to try out Whisper, so here we are.
Feel free to open a Push/Pull Request if you want to add something.
I'm sure you can figure it out by reading the code for now.
I'll add a more detailed explanation later, if this turns out to be useful.