v1.4.5 runtime-llamacpp-v0.2.1
modelscope/FunASRAug 27, 2026 → Aug 26, 20260 days between releases
Generating AI comparison
Analyzing what changed between v1.4.5 and runtime-llamacpp-v0.2.1. This page updates automatically — usually within a minute.
The Two Releases
FunASR 1.4.5
Aug 27, 2026· 2 months ago
This release simplifies the default Python inference installation by removing `torchaudio` as a hard dependency and introducing a `kaldi-native-fbank` backend. It also bundles prebuilt llama.cpp runtimes alongside the Python distributions to streamline deployment.
Full release notesFunASR llama.cpp runtime v0.2.1
Aug 26, 2026· 2 months ago
Introduces prebuilt self-contained binaries for the FunASR llama.cpp runtime, supporting SenseVoice, Paraformer, and Fun-ASR-Nano models with built-in FSMN-VAD. It offers various hardware optimizations including AVX2, Vulkan, and CUDA backends to enhance performance across different platforms.
Full release notes