Skip to content

Dataset · ShoukanLabs

Anime voices with text captions for TTS

ShoukanLabs/AniSpeech

A collection of voiced anime audio with text captions for training speech synthesis.

What's inside

The dataset contains audio recordings of anime voices with text captions. The collection is continuously updated, with newly annotated audio added and organized by language. The language is English, and the volume ranges from 10k to 100k examples. A major update is also announced in the card.

What you can do with it

  • Train a TTS model on anime voices
  • Fine-tune LoRA for character-style speech synthesis
  • Test voice cloning on captioned audio
Tasks
Text → speech
Languages
English
Size
10K–100K rows
Downloads this month
712
Likes
69

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("ShoukanLabs/AniSpeech")