The llama.cpp project has released version b9950, which includes code quality improvements such as adding a unit test for llama-batch and fixing Windows 32-bit build issues.

  • Added unit test for llama-batch functionality.
  • Fixed win32 builds.
  • Added not implemented assertions in unused methods.
  • Removed unreachable code.

This release provides updated binaries for macOS, Linux, Android, Windows, and openEuler across various hardware backends including CPU, CUDA, Vulkan, ROCm, OpenVINO, SYCL, and HIP.