YT-Data/ja-en-syn-qwen3-mimi

Name: YT-Data/ja-en-syn-qwen3-mimi
Creator: YT-Data
License: apache-2.0
Keywords: huggingface, task_categories:text-to-speech, language:en, language:ja, license:apache-2.0, size_categories:1M<n<10M, region:us, text-to-speech

Text To Speech · YT-Data· 821

apache-2.0 15 GB

YT-Data/ja-en-syn encoded for OrpheusPlus training: Qwen3-4B text tokens + Mimi audio tokens in the Orpheus prompt layout. Single column inputids. Built by encodemimi.py.

Open in MLForge Sign up free Desktop app

# download instantly
mlforge datasets pull YT-Data/ja-en-syn-qwen3-mimi

Dataset details

Task

Text To Speech

Language

License

apache-2.0

Size

15 GB

Rows / images

6.0M

Creator

YT-Data

Downloads

821

Source

huggingface_datasets

Updated

2026-05-24

About YT-Data/ja-en-syn-qwen3-mimi

YT-Data/ja-en-syn encoded for OrpheusPlus training: Qwen3-4B text tokens + Mimi audio tokens in the Orpheus prompt layout. Single column inputids. Built by encodemimi.py.