llama.cpp 发布: b11193

llama.cpp 发布 · 来源更新 2026/09/26 11:33

今日摘要使用自己的 API，仅供个人查看

来源摘要

hexagon: find software divide calls using binary inspection tool (#29449) * hex-scripts: fix table alignment * hex-scripts: find sw div calls using binary inspection tool **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-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/b11193/llama-b11193-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-arm64.tar.gz) - [Ubuntu s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-s390x.tar.gz) - [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-vulkan-x64.tar.gz) - [Ubuntu arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-vulkan-arm64.tar.gz) - [Ubuntu x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-cuda-12.8-x64.tar.gz) - [CUDA 12.8 libraries](https://github.com/ggml-org/llama.cpp/releases/download/b11193/cudart-llama-b11193-bin-ubuntu-cuda-12.8-x64.tar.gz) - [Ubuntu x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b11193/llama-b11193-bin-ubuntu-cuda-13.4-x64.tar.gz) - [CUDA 13.4 libraries](https://github.com/ggml-org/ll

阅读原始来源

来源: llama.cpp 发布 · 官方来源
来源发布: 2026/09/26 11:30
来源更新: 2026/09/26 11:33
首次采集: 2026/09/26 11:59

本文为公开信息索引与摘要，详情及后续变化请以原始来源为准。

Meta · Llama 动态 NVIDIA · 英伟达动态 Intel · 英特尔动态

llama.cpp 发布: b11192

vendor : update cpp-httplib to 0.58.0 (#29407) **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11192/llama-b11192-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, K

来源更新 2026/09/26 08:28国际查看原文

算力与芯片IT之家 AI 与硬件媒体报道

Anthropic 与 Akamai 达成 7 年 116 亿美元协议，扩充 CPU 算力

IT之家 9 月 26 日消息，云计算、网络安全、内容交付企业 Akamai 当地时间 24 日宣布大幅扩展与 Anthropic 的合作关系，两家公司签署了一份为期 7 年、价值 116 亿美元（IT之家注：现汇率约合 779.2 亿元人民币）的合同。 ▲ Akamai 总部根据协议，Akamai 将以其分布式人工智能基础设施和软件满足 Anthropic 日益增长的 CPU 工作负载需求。这 116 亿美元承诺对应 Akamai 的 55 亿美元额外总资本支出。双方还可将合同扩大最多 90 亿美元。作为两家

发布 2026/09/26 07:00国内查看原文

算力与芯片llama.cpp 发布官方来源

llama.cpp 发布: b11191

common,rpc : simplify fs_create_directory_with_parents() (#29432) The original function was broken on Windows for some unicode paths Paths without a trailing separator now create the last directory too, matching the function name. All current callers already i

来源更新 2026/09/26 04:30国际查看原文

算力与芯片llama.cpp 发布官方来源

llama.cpp 发布: b11190

mtmd: fix mel preprocessor in LFM2 audio (#29403) which resulted in different greedy transcripts for 4.5% of English and 6.5% of Japanese test utterances. In Japanese, some differences changed entire words. This change: * uses `log(x + 2^-24)` instead of clamp

来源更新 2026/09/26 04:02国际查看原文

算力与芯片llama.cpp 发布官方来源

llama.cpp 发布: b11189

opencl: add bin kernel `kernel_gemm_noshuffle_q5_k_f32_32b_trans_ila_a8_bin`, `kernel_gemm_noshuffle_q5_k_q8_1_dp4a_ila_a8_bin` (#29401) * opencl: add A8 Q5_K non-MoE non dp4a + dp4a binary kernel * opencl: fix s transpose - s only transposed for bin kernels -

来源更新 2026/09/26 03:37国际查看原文