You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Fine-tuning toolkit for Maya1, a 3B-parameter open-source TTS model by Maya Research. Supports full fine-tuning with SNAC neural codec, natural language voice design, and inline emotion tags. Features YAML-based config, offline preprocessing, and automatic audio sample generation at every checkpoint.