tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-26 19:55:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

CUDA: refactor mmq, dmmv, mmvq (#7716)

Browse Source 
* CUDA: refactor mmq, dmmv, mmvq

* fix out-of-bounds write

* struct for qk, qr, qi

* fix cmake build

* mmq_type_traits
This commit is contained in:
Johannes Gäßler
2024-06-05 16:53:00 +02:00
committed by GitHub
parent 2b3389677a
commit 7d1a378b8f
112 changed files with 1783 additions and 1767 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
Makefile
View File

@ -444,6 +444,7 @@ ifdef LLAMA_CUBLAS
endif
OBJS_CUDA_TEMP_INST = $(patsubst %.cu,%.o,$(wildcard ggml-cuda/template-instances/fattn-wmma*.cu))
OBJS_CUDA_TEMP_INST += $(patsubst %.cu,%.o,$(wildcard ggml-cuda/template-instances/mmq*.cu))
ifdef LLAMA_CUDA_FA_ALL_QUANTS
OBJS_CUDA_TEMP_INST += $(patsubst %.cu,%.o,$(wildcard ggml-cuda/template-instances/fattn-vec*.cu))
else