2 days ago
-
Fix _wave() under autograd
-
README: link to the in-browser demo
3 days ago
-
v2.3: fine-tuned BigVGAN, Vocos, fast mode, voice cloning, number reading, long texts
4 days ago
-
README: hardware, speed and memory; fast mode
-
v2.2: speaking styles (reading, conversational), 592 voices, new README and samples; CC BY-SA 4.0 (DailyTalk)
-
Shorter gaps between sentences (sentence_gap 0.35 s) and app.py, a local page to try your own text
5 days ago
-
Studio voice as default, timbre matching, clip finishing; model card: One steady voice
-
Frontend docstring
-
mmmhusto TTS English: v2.1 (42 M acoustic + 14 M BigVGAN base), code, model card, samples
-
initial commit