Repo of the code from the Medium article
This is what comes out from this repo
this code is tested on windows, with python 3.10
Create a virtual environment and install the following packages
pip install whisper-cpp-pybind #good for pytho 3.10
pip install python-ffmpeg
pip install streamlit==1.26.0
You also need to install FFmpeg, and ad it to path
winget install "FFmpeg (Essentials Build)"
or download it and install as described in official website https://ffmpeg.org/download.html
Download the models already quantized from the official Hugging Face repo of ggerganov
https://huggingface.co/ggerganov/whisper.cpp/tree/main
The models come in different sizes, and quantization methods
download them into a subfolder called models
for this app i used ggml-tiny.en-q8_0.bin
from the terminal, with venv
activated run
streamlit run stWhisper.py