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

Merged

Commits

Add mul_mm_f16_f32_kq_kqv kernel

shaofeiqi committed 47 days ago
Add ggml_cl_mul_mat_kq_kqv_adreno func

shaofeiqi committed 47 days ago
fix whitespace

shaofeiqi committed 47 days ago
remove unused variable

shaofeiqi committed 47 days ago
remove redundant

shaofeiqi committed 47 days ago
refactor and clean up

shaofeiqi committed 43 days ago
remove trailing whitespace

shaofeiqi committed 41 days ago

Loading