llama.cpp 发布: b11168

来源摘要

hexagon: dynamic quantizer improvements (#29395) * hexagon: fix accuracy issue in Q8_0 N=1 MUL_MAT * hex-quant: fix register spills * hex-mm: use dma for all dyn.quant paths Co-authored-by: Aparna M P * hex-mm: remove obsolete run_quant_task * hex-mm: update tracing to properly wrap the events * hex-mm: use act for activation data in all paths * hex-mm: use act_ instead of src1_ to avoid confusion in fused kernels * hex-mm: remove/reroute the rest of the non-DMA act (aka src1) logic * hex-dma64: yet another pass at cleaning up the dma_addr_t casts * Update ggml/src/ggml-hexagon/htp/matmul-ops.h Co-authored-by: Sigbjørn Skjæret * Update ggml/src/ggml-hexagon/htp/matmul-ops.c Co-authored-by: Sigbjørn Skjæret * Update ggml/src/ggml-hexagon/htp/matmul-ops.c Co-authored-by: Sigbjørn Skjæret * Update ggml/src/ggml-hexagon/htp/matmul-ops.c Co-authored-by: Sigbjørn Skjæret --------- Co-authored-by: Max Krasnyansky Co-authored-by: Aparna M P Co-authored-by: Sigbjørn Skjæret **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11168/llama-b11168-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b11168/llama-b11168-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11168/llama-b11168-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11168/llama-b11168-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11168/llama-b11168-bin-ubuntu-arm64.tar.gz) - [Ubuntu s390x (CPU)](https://github.com