MMLForge Search models, datasets, tasks… Start free
Home › Datasets › FDlalala/tranS
T

FDlalala/tranS

Audio Classification · FDlalala· 2.5K
cc-by-nc-4.0 161 GB

DeepDialogue-xtts is a large-scale multimodal dataset containing 40,150 high-quality multi-turn dialogues spanning 41 domains and incorporating 20 distinct emotions with coherent emotional progressions. This repository contains the XTTS-v2 variant of the dataset, where speech is generated using XTTS-v2 with explicit emotional conditioning.

Open in MLForge Sign up free Desktop app
# download instantly
mlforge datasets pull FDlalala/tranS

Dataset details

Task
Audio Classification
Language
en
License
cc-by-nc-4.0
Size
161 GB
Rows / images
243.3K
Creator
FDlalala
Downloads
2.5K
Source
huggingface_datasets
Updated
2026-05-28

About FDlalala/tranS

DeepDialogue-xtts is a large-scale multimodal dataset containing 40,150 high-quality multi-turn dialogues spanning 41 domains and incorporating 20 distinct emotions with coherent emotional progressions. This repository contains the XTTS-v2 variant of the dataset, where speech is generated using XTTS-v2 with explicit emotional conditioning.