llama.cpp
vulkan: in graph_optimize, try to group ADD operations
#18060
Merged

Loading