6d ago · llama.cpp releases →
webgpu : add COUNT_EQUAL support …
Includes: llama.cpp b12199, b12200: opencl: improve q8_0 and MoE small-n verification for spec (#31831), llama.cpp b12201, llama.cpp b12202, b12203: ggml : switch to F32 src1 in the CPU backend (#31854), b12204: ci: move runners to Ubuntu 26.04 (#31529), b12205: webgpu : add COUNT_EQUAL support (#31578)