runtime-llamacpp-v0.1.5
modelscope/FunASRruntime-llamacpp-v0.1.5Jul 13, 2026by github-actions[bot]
AI Summary
Introduces prebuilt self-contained binaries for the llama.cpp runtime, enabling on-device ASR for SenseVoice, Paraformer, and Fun-ASR-Nano without requiring Python or build steps.
Key Highlights
- Provides prebuilt self-contained binaries for llama.cpp runtime
- Supports SenseVoice, Paraformer, and Fun-ASR-Nano models
- Includes built-in FSMN-VAD for Whisper.cpp-style on-device processing
- No Python dependency or compilation required
New Features
- Prebuilt runtime binaries for x64 and x64-avx2
- CLI tools: llama-funasr-cli, llama-funasr-sensevoice, llama-funasr-paraformer
- Integrated FSMN-VAD functionality
Full Release Notes
Prebuilt self-contained binaries for the FunASR llama.cpp / GGUF runtime — SenseVoice, Paraformer and Fun-ASR-Nano with built-in FSMN-VAD (a whisper.cpp-style on-device ASR, strong on Chinese). Get a model with `bash download-funasr-model.sh <sensevoice|paraformer|nano>`, then run `llama-funasr-cli` / `llama-funasr-sensevoice` / `llama-funasr-paraformer`. Use the default x64 asset for maximum CPU compatibility; use the x64-avx2 asset on CPUs with AVX2/FMA/F16C/BMI2 for higher throughput. No Python, no build. Docs: runtime/llama.cpp/README.md