mirror of
https://github.com/likelovewant/ollama-for-amd.git
synced 2025-12-21 14:26:30 +00:00
clang outputs are faster. we were previously building with clang via gcc wrapper in cgo but this was missed during the build updates so there was a drop in performance