Skip to content

Dataset · simon3000

Genshin Impact voice lines

simon3000/genshin-voice

654k wav files of Genshin Impact character voice lines in Chinese, English, Japanese, and Korean.

What's inside

The dataset contains voice lines from the game Genshin Impact: 654,252 wav files. Some records lack a speaker (7,291, 1%), a transcription (52,693, 8%), and an inGameFilename (1,088, 0%). Languages include Chinese, English, Japanese, and Korean; downloads by speaker are grouped by language and ZIP size.

What you can do with it

  • Train ASR for game voice lines
  • Fine-tune TTS or LoRA for character voices
  • Audio-based speaker classification
Tasks
Audio classification, Automatic speech recognition, Text → speech
Languages
Chinese, English, Japanese, Korean
Size
100K–1M rows
Downloads this month
18.6K
Likes
307

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("simon3000/genshin-voice")