neriqlabs/kasanoma-tts-twi-v0.5
Text-to-Speech • Updated • 24
AI & ML interests
We build the data AI cannot reach, and the models that prove it. Underrepresented languages, cultures and communities, in whatever modality the internet never captured. Starting with Asante Twi: open speech recognition (TubaSTT), speech synthesis (Kasanoma) and a small tool-calling LLM. Twi, Ga, Ewe, Dagbani, Hausa.
Recent Activity
View all activity