Skip to content

Dataset · joujiboi

Voice acting lines from Japanese visual novels

joujiboi/Galgame-VisualNovel-Reupload

A multi-million corpus of Japanese game voice acting and texts repacked for easy downloading.

What's inside

The repository is a reorganized version of a well-known dataset of voice acting from visual novels in Japanese. Inside are 1 to 10 million pairs of audio tracks with character lines and text transcripts, combined into a single train split for streaming download via the datasets library without manual archive unpacking and JSON parsing.

What you can do with it

  • Training Japanese speech recognition (ASR) models on emotional lines
  • Training expressive speech synthesis (TTS) systems with diverse voices
  • Creating voice models and LoRA for voicing anime and game characters
Tasks
Automatic speech recognition, Text → speech
Languages
Japanese
Size
1M–10M rows
Downloads this month
4.3K
Likes
42

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("joujiboi/Galgame-VisualNovel-Reupload")