LibriTTS-R Metadata Generator

Processes the LibriTTS-R speech synthesis dataset and generates a SQLite database schema with books, chapters, speakers, and transcriptions.

Quick Start

./run.sh

What it does

• Downloads LibriTTS-R dataset from OpenSLR
• Parses transcription files and metadata
• Generates SQLite database schema
• Converts WAV files to MP3 format

Requirements

• Python 3.10+
• FFmpeg
• wget
• sqlite3 (for local testing)

Output

• dist/01_schema.sql - Database schema and initial data (books, speakers, chapters)
• dist/02_transcriptions_*.sql - Transcription data split into chunks of 650 records
• dist/03_indexes.sql - Database indexes for performance optimization
• dist/ - MP3 audio files (128kbps)

Local Database Testing

You can test the generated SQL files locally using sqlite3:

# Import all SQL files into a local database
cat dist/01_schema.sql dist/02_transcriptions_*.sql dist/03_indexes.sql | sqlite3 libritts-r.db

# Verify the data
sqlite3 libritts-r.db "SELECT COUNT(*) FROM transcriptions; SELECT COUNT(*) FROM books; SELECT COUNT(*) FROM chapters; SELECT COUNT(*) FROM speakers;"

Cloudflare Deployment

To deploy to Cloudflare D1 (database) and R2 (audio storage):

./publish.sh

This script will:

1. Create D1 database libritts-r if it doesn't exist
2. Execute all SQL files to populate the database with metadata
3. Create R2 bucket libritts-r if it doesn't exist
4. Upload all MP3 files to R2, preserving the directory structure

The deployment creates a complete cloud setup with:

• Structured metadata in D1 (books, chapters, speakers, transcriptions)
• Audio files in R2 (MP3 files organized by dataset subset and chapter)

License

This project is licensed under the MIT License. See the LICENSE file for details.