Multilingual NLP: Transcription and Translation

Setup Notes

Important: For audio processing (mp3/m4a conversion and transcription), use Python 3.12 or lower. Python 3.13+ is not supported due to removal of the audioop module, which is required by pydub and other audio libraries.

Recommended: Install Python 3.12 and create your virtual environment with it. Gen AI Mini Projects

Note for macOS Users

If you encounter errors when installing PyAudio (such as portaudio.h file not found), you need to install the PortAudio library using Homebrew:

brew install portaudio

After installing PortAudio, re-run:

pip install -r requirements.txt

This will allow PyAudio to build and install successfully on macOS.

Running the Apps

Note: When running the Streamlit app for the first time, it may take a while to start because it needs to download about 200MB of data before loading. Please be patient during the initial startup.

How to Run the Apps

• To run the live transcription app (using your microphone):

  cd Live_audio2text_app
  streamlit run app.py

• To run the audio-to-text app for MP3/M4A files:

  cd audio2text_from_mp3
  streamlit run streamlit_app.py