Skip to content

Latest commit

 

History

History
62 lines (46 loc) · 1.48 KB

README.md

File metadata and controls

62 lines (46 loc) · 1.48 KB
Raw

Loquitur

Loquitur is a web application that allows users to upload videos or download them from BigBlueButton. It enables users to view transcriptions of these videos and interact with them using the Ollama AI tool for asking questions about the video content.

Requirements

  • Python 3.10
  • WhisperX
  • NVM or Node 20
  • Ollama (Optional, only for AI chat)

Installation

First, ensure that you have requirements installed in your environment. Then, follow these steps to install Loquitur:

# Install moviepy, a dependency for video processing. Use the same virtualenv as WhisperX
pip install moviepy==1.0.3

# Copy the sample environment file and configure it
cp .env.sample .env
# Remember to edit the .env file with your specific settings

# Install Node.js and npm dependencies
nvm install
npm install

Running the Application

To run Loquitur, start both the frontend and the API server:

# Start the frontend interface
npm start

# Start the API server
npm run start:api

Upcoming Features (TODO)

  • Implement chat shortcuts for ease of use.
  • Add functionality to download videos with subtitles.
  • Enable audio file uploads.
  • Upload with skeleton data & persitance on reload.
  • Upload & AI error handling.

Technology Stack

  • Nx
  • Angular 17
  • Ollama
  • WhisperX
  • Fastify
  • Open props
  • TRPC
  • Zod
  • Tanstack
  • Fuse.js
  • RxAngular
  • AngularMaterial