mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-07-16 01:15:58 +02:00
94242a62c0
While buffers reside in host memory, additional transformation is needed to use buffers with zDNN. Fixes #18848