mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-09-18 00:34:55 +02:00
4c9233c034
* cuda : enable i16 and i32 for DUP * docs : update ops table for DUP on CUDA