●Change8

b9004

📦 llama-cppView on GitHub →
✨ 6 features

Summary

This release synchronizes updates with ggml and provides a comprehensive set of pre-built binaries for macOS, Linux, Android, Windows, and openEuler targeting various CPU, GPU, and specialized hardware backends.

✨ New Features

  • Synchronized updates to ggml.
  • Provided pre-built binaries for macOS (Apple Silicon and Intel) with optional KleidiAI support.
  • Provided pre-built binaries for Linux across various architectures (x64, arm64, s390x) supporting CPU, Vulkan, ROCm 7.2, OpenVINO 2026.0, and SYCL (FP32/FP16).
  • Provided pre-built binaries for Android (arm64 CPU).
  • Provided pre-built binaries for Windows supporting CPU, CUDA 12.4/13.1, Vulkan, SYCL, and HIP (Radeon).
  • Provided pre-built binaries for openEuler supporting various architectures and accelerators (310p, 910b with ACL Graph).