mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-06-29 15:50:22 +00:00
94242a62c0
While buffers reside in host memory, additional transformation is needed to use buffers with zDNN. Fixes #18848