Latest / Kana & Mari's SoundRepos (English)

n4ze3m/tiny-voca | compact two voice English text to speech | GitHub
tiny-voca is a compact English text-to-speech model with two American English voices, designed to run on CPUs and in browsers. It includes inference, training, data preparation, evaluation, and ONNX export tools, with a source-filter waveform generator and adjustable speaking speed and pitch. Related: n4ze3m/tiny-voca / compact two voice English text to speech / GitHub
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.