Dataset · joujiboi
Japanese anime speech with transcripts
joujiboi/japanese-anime-speech
Thousands of Japanese anime speech audio fragments from visual novels with transcripts for ASR.
What's inside
An audio-text dataset for training automatic speech recognition. It contains thousands of audio clips and corresponding transcriptions from various visual novels in Japanese. The goal is to improve the accuracy of ASR models like Whisper on dialogues from anime and similar Japanese media.
What you can do with it
- Fine-tuning Whisper for Japanese anime speech
- Evaluating ASR on dialogues from visual novels
- Creating a LoRA/adapter for anime speech recognition
- Tasks
- Automatic speech recognition
- Languages
- Japanese
- Size
- 10K–100K rows
- Downloads this month
- 877
- Likes
- 166
Models trained on it
- nvidia/magpie_tts_multilingual_357mText → speech 16.5K
- mradermacher/Anime-Llasa-3B-i1-GGUF 814
- mradermacher/Anime-Llasa-3B-GGUF 480
- NandemoGHS/Anime-Llasa-3BText → speech 479
- mradermacher/Galgame-Llasa-3B-v3-GGUF 233
- mradermacher/Galgame-Llasa-1B-v3-GGUF 187
- mradermacher/Galgame-Llasa-1B-v2-GGUF 168
- Ananie/japanes-ttsText → audio 48
How to load it
from datasets import load_dataset
ds = load_dataset("joujiboi/japanese-anime-speech")