llama.cpp
748d4225 - ggml-cuda: HIP: optimize packed byte subtraction (`__vsubss4` -> `__vsub4`) (#29478)

Commit
50 days ago
ggml-cuda: HIP: optimize packed byte subtraction (`__vsubss4` -> `__vsub4`) (#29478) * ggml-cuda: HIP: optimize non-saturating packed byte subtraction (`__vsubss4`) * CI: ignore 1 spilled vgpr in fattn_vec --------- Co-authored-by: Carl Philipp Klemm <carl@uvos.xyz>
Parents
Loading