Add windows build scripts for CUDA (#924)

Fix block-size assumption (#935)

fix broken link 'feel free to choose' -> example llamafiles (#927)

Fix: GGUF Q5_1 quant crashes llamafile on aarch64 cpu (#928)

Add tinyblasStrsmBatched kernel (#923)

Add support for vulkan dylibs (#892)

Fix README_0.10.0 (#918)

llamafile reloaded (v0.10.0) (#867)

update supported OpenBSD versions (#897)

Fixed broken llamafile URL in docs (#873)

fix(update-llama-cpp): Use `new_build_wip` as base ref. (#850)

chore: Add `update-llama-cpp` workflow. (#846)

feat: move docs to mkdocs (#824)

chore: llama.cpp as submodule (#819)

chore: convert stable diffusion to submodule (#818)