List your company

Text-to-Speech on Seed-TTS hard (test)

7.53WERSemantic-VAE
3.938828.179452.4276.6606May 28, 2026

Evaluation results

MethodLinks
Semantic-VAEZero-shot=true2026.057.5367
HoliTokZero-shot=true2026.057.5966
MingTok-AudioZero-shot=true2026.0514.7566
HoliTok-UniteModeling=Unified spoken language modeling2026.0516.7962
HoliTok-BaseModeling=Unified spoken language modeling2026.0530.4461
MingTok-AudioModeling=Unified spoken language modeling2026.0550.3555
Semantic-VAEModeling=Unified spoken language modeling2026.0597.3161