⬢github Python · 786 ★ +26 since we first saw it · pushed 10 h ago · Apache-2.0
n4ze3m/tiny-voca
A 4.5M-parameter two-voice English text-to-speech model, trained from scratch
tiny-voca is a tiny open-source English text-to-speech model with 4.5M parameters (18 MB) and two fixed voices (male and female). It runs fast on CPU or in the browser via PyTorch, ONNX Runtime, or plain JavaScript, and ships with code, weights, dataset, and full training recipe — trainable in about 8 hours on one GPU.
Why now: It's trending as one of the most-starred new repos, likely thanks to its tiny size, browser demo, and a growing community porting it to German, Italian, Turkish, and Korean.
Who it is for: Developers who want lightweight, fast, open TTS that runs locally or in the browser without a GPU or cloud API.
onnxtext-to-speechtts
Stars over our 105 snapshots: 760 to 786, since 1 d ago.
Where people talked about it
- ⬢github new repos, most starred 14 min ago
- ⬢github new repos, most starred 17 h ago
API: https://socialmediatrends-api.osmike.com/v1/repos/n4ze3m/tiny-voca