vibevoice.cpp — quantized model bundle Brought to you by the LocalAI team — the creators of LocalAI, the open source AI engine that runs any model — LLMs, vision, voice, image, video — on any hardware. No GPU required. Quantized GGUF weights for vibevoice.cpp, a C/C++ port of Microsoft VibeVoice (TTS + ASR) on top of ggml . File Source Quant Size vibevoice realtime 0.5B q8 0.gguf microsoft/VibeVoice Realtime 0.5B Q8 0 (matmul) + F16 ~1.6 GB vibevoice asr q8 0.gguf microsoft/VibeVoice ASR Q8 0 (matmul) + F16 ~13 GB voice en Carter man.gguf upstream voice prompt cache F16 8 MB voice en Emma.gguf upstream voice prompt cache F16 6 MB tokenizer.gguf Qwen2.5 BPE + VibeVoice specials — 6 MB Quantization scheme scripts/quantize gguf.py in the source repo selectively quantizes only the LM matmul weights — attention q/k/v/o, ffn gate/up/down, and lm head — to Q8 0. Everything else (1 D conv kernels, RMSNorm scales, biases, layer scale gammas, token embeddings, small scalars) passes through unchanged. The conv1d implementation in vibevoice.cpp casts kernels to F16 inline rather than dequantizing on the fly, so quantizing those would corrupt the convolution outputs. Q8 0 was chosen because it'…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy