Transcribing

To transcribe the audio files recorded from the participant, you first need to create a basic virtual environment due to some issues using the one that comes with this software.

python -m venv .venv
source .venv/bin/activate
pip install -U openai-whisper

The first time you run the script it will have to download the model, which takes a few minutes.

To run the script:

python rtcog/matching/transcribe.py -i <input_dir> -o <output_dir> -p <input_prefix> -m <model>