inference-software

b11715

Observed ; updated .

ggml-cuda: HIP: use __builtin_amdgcn_perm for q1_0 unpacking (#29927) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11715/llama-b11715-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI…

Sources

Reference: Llama

Why it matters
Not yet written for this story.
Who is affected
Not yet written for this story.