Search is not available for this dataset
audio audioduration (s) 5.19 10 | label class label 16
classes |
|---|---|
0cosyvoice2_Angry | |
0cosyvoice2_Angry | |
1cosyvoice2_Happy | |
1cosyvoice2_Happy | |
2cosyvoice2_Neutral | |
2cosyvoice2_Neutral | |
3cosyvoice2_Sad | |
3cosyvoice2_Sad | |
4cosyvoice2_Surprise | |
4cosyvoice2_Surprise | |
5cosyvoice3_Angry | |
5cosyvoice3_Angry | |
6cosyvoice3_Happy | |
6cosyvoice3_Happy | |
7cosyvoice3_Neutral | |
7cosyvoice3_Neutral | |
8cosyvoice3_Sad | |
8cosyvoice3_Sad | |
9cosyvoice3_Surprise | |
9cosyvoice3_Surprise | |
10doctor5_clone_source | |
10doctor5_clone_source | |
10doctor5_clone_source | |
10doctor5_clone_source | |
10doctor5_clone_source | |
11indextts25_Angry | |
11indextts25_Angry | |
12indextts25_Happy | |
12indextts25_Happy | |
13indextts25_Neutral | |
13indextts25_Neutral | |
14indextts25_Sad | |
14indextts25_Sad | |
15indextts25_Surprise | |
15indextts25_Surprise |
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
TTS-Emo-Demo (20260913)
Emotion-effect comparison across TTS models on the SAME fixed text.
What's here
- Fixed text: the opening doctor turn + patient reply of a warmth-5 showcase dialogue (rheumatology follow-up, RA patient; text in each folder's meta.json).
- Each
<tts_model>_<emotion>/folder: the same two turns synthesized under that emotion, wav (<=10 s each, trimmed), meta.json (emotion, model, text). - Models so far: FunAudioLLM/CosyVoice2-0.5B (cosyvoice2) and FunAudioLLM/Fun-CosyVoice3-0.5B-2512 (cosyvoice3). Same seeds, same text, same emotion instruction map - direct A/B listening.
Emotions (5)
Neutral, Happy, Sad, Angry, Surprise.
Scope: research demo, synthetic clinical-roleplay content only. Generated 2026-09-13 by the SIMAX v2 pluggable TTS harness.
- Downloads last month
- 71