Skip to content

Latest commit

 

History

History

notebooklm

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

NotebookLM - Self-hosted Audiobook Generation Platform

Music from Book
Voices from Book

Overview

NotebookLM is a self-hosted platform designed to create, manage, and enhance audiobooks. This platform leverages various tools and models to parse, transcribe, and improve manuscripts, ultimately providing high-quality audio content.

"Audiobook Generator"

Steps

  • python audiobook.py
    • converts script pdf into structured scene json using mistral-large API.
    • triggers Text to Speech server for speech/dialogs and store audio files in resources folder
    • triggers Text to Sound server for background sound/music and store audio files in resources folder
    • Combines generated audio from speech and sound modules and creates audibook for the provided script

"Script Parser"

  • Prompts for Audiobook: Creating a storyline using prompts Read more

Current Status

  1. Script parser - convert to scene structured json
  2. TTS server - Parler-tts server to generate speech for dialogs in scenes
  3. AudioGen module - Audiocraft/magnet background sound creator
  4. Basic audiobook - linear worflow for full audio creation without scene logic

Documentation

  • Challenge: project challenge. Read more
  • Approach: Approach in solving the problem. Read more
  • Workflow: Worflow for audiobook generation. Read more
  • Setup: Follow the setup guide to build an End-to-End platform for Audibook Generation/Notebook LLM.
  • UX: UX design for Audiobook generation Read more
  • Database Schema: Schema for Scene Editing and Media content Read more

An experiment to build a production grade audiobook content generator system to help publishers build on their IP and reach larger audience.

Reference