opencl: add kernel to handle mat mul in attention to improve encoding speed#17181

max-krasnyansky merged 7 commits intoggml-org:masterggml-org/llama.cpp:masterfrom

qualcomm:sq/mul-mat-kq-kqv-opt-newqualcomm/llama.cpp:sq/mul-mat-kq-kqv-opt-newCopy head branch name to clipboard