AbbieSpeak

Saved voice state for Kyutai Pocket TTS, made with pocket-tts export-voice. Nothing is trained: the file is loaded and only the generation step runs.

pocket-tts generate --voice hf://theabbie/AbbieSpeak/abbie.safetensors --text "Hello there."

Or with no setup besides uv (add HF_TOKEN while the repo is private):

hf download theabbie/AbbieSpeak speak.py --local-dir .
uv run speak.py "The quick brown fox jumps over the lazy dog." out.wav

Generation needs only the open kyutai/pocket-tts-without-voice-cloning weights; no gated access is required.