SynDataLab-JA/irodori-clones-3m-v2-no-emoji

Name: SynDataLab-JA/irodori-clones-3m-v2-no-emoji
Creator: SynDataLab-JA
License: apache-2.0
Keywords: huggingface, task_categories:text-to-speech, language:ja, license:apache-2.0, size_categories:1M<n<10M, format:parquet, modality:audio, modality:text, library:datasets, text-to-speech

Text To Speech · SynDataLab-JA· 9.4K

apache-2.0 2.2 TB

3,290,000 cloned utterances generated with Aratako/Irodori-TTS-500M-v2, using the 10,000 reference voices from SynData-2/irodori-refs-10k-v2. 329 clones per ref voice, each with a unique Japanese conversational text.

Open in MLForge Sign up free Desktop app

# download instantly
mlforge datasets pull SynDataLab-JA/irodori-clones-3m-v2-no-emoji

Dataset details

Task

Text To Speech

Language

License

apache-2.0

Size

2.2 TB

Rows / images

3.3M

Classes

Creator

SynDataLab-JA

Downloads

9.4K

Source

huggingface_datasets

Updated

2026-06-01

SynDataLab-JA/irodori-clones-3m-v2-no-emoji

Dataset details

About SynDataLab-JA/irodori-clones-3m-v2-no-emoji