Speech-to-Text Service

A Speech-to-text transcription service using AssemblyAI.

Setup

• Python version: Python 3.11.4
• Virtual environment (pyenv): pyenv 2.3.25
• Poetry: 1.6.1

If you have PyEnv installed, you may create a virtual environment from the project repo,

$ pyenv virtualenv 3.11.4 "<your-venv-name>"
$ pyenv local "<your-venv-name>"

To install dependencies using Poetry,

$ poetry install

Additional dependencies from AssemblyAI,

# (Mac)
$ brew install portaudio

# (Debian/Ubuntu)
$ apt install portaudio19-dev

To avoid exposing of your API key, add it as an environment variable/secret,

$ export AAI_API_KEY="<your-API-key>"

Usage

To transcribe an audio file, run the script below

$ python modules/transcribe_audio_file.py -af "https://github.com/AssemblyAI-Examples/audio-examples/raw/main/20230607_me_canadian_wildfires.mp3"

To transcribe audio from a microphone, run the script below

# Press Ctrl+C to stop recording
$ python modules/transcribe_mic_audio.py