Skip to content

Latest commit

 

History

History
69 lines (59 loc) · 2.34 KB

README.md

File metadata and controls

69 lines (59 loc) · 2.34 KB

Whisper Transcribe

This project uses OpenAI's Whisper-tiny to transcribe .mp3 files, store transcriptions and search for existing transcription text in the database.

Quick Start

Windows

  • Run start.bat

Linux

  • Run start.sh

Run tests

ng test - run Angular tests test_app.bat or test_app.sh - run Flask tests

Tests

Frontend

  1. POST "/transcribe" uploadTranscribe() success check - being able to transcribe something is the most important part of the whole app
  2. POST "/transcribe" uploadTranscribe() failure check - making sure the error handling for the most important part of the app works
  3. html element rendering - every view uses the same method of displaying the fetched data, so it should at least be displaying correctly

Backend

  1. Transcribe __init__ test - main entity class must work
  2. GetAllTransactions query - even if uploads break at least previously processed data can be seen.
  3. transcribeAndSave() unit test

Stack

Backend

  • Flask-RESTful
  • SQLite

Frontend

  • Angular

Commands used to set this up

Backend (Flask)

python -m venv venv
venv\Scripts\activate
pip install Flask
pip install Flask-RESTful
pip install sqlalchemy
pip install transformers torch torchvision torchaudio
pip install flask-cors
pip freeze > requirements.txt

Frontend (Angular)

npm install -g @angular/cli
ng new frontend
> CSS
> no Server Side Rendering or Static Site Generation
ng g config karma

Dev References