matt/ollama37
1
0
Fork 0
mirror of https://github.com/dogkeeper886/ollama37.git synced 2025-12-20 20:57:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama: update vendored code to commit 40c6d79f (#7875)

Browse Source
This commit is contained in:
Jeffrey Morgan
2024-12-10 19:21:34 -08:00
committed by GitHub
parent a37f4a86a7
commit 527cc97899
289 changed files with 58552 additions and 41806 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
make/Makefile.rocm
View File

@@ -51,7 +51,7 @@ GPU_DIST_LIB_DEPS= $(sort $(addprefix $(ROCM_DIST_DEPS_DIR)/,$(notdir $(GPU_LIBS
ROCBLAS_DIST_DEP_MANIFEST = $(ROCM_DIST_DEPS_DIR)/rocblas/library/TensileManifest.txt
ifeq ($(OS),linux)
GPU_COMPILER_FPIC := -fPIC -Wno-unused-function -std=gnu++11
GPU_COMPILER_FPIC := -fPIC -Wno-unused-function -std=gnu++17
else ifeq ($(OS),windows)
GPU_COMPILER_FPIC := -Xclang --dependent-lib=msvcrt
endif
@@ -69,11 +69,13 @@ GPU_COMPILER_CUFLAGS = \
-O3 \
-DGGML_USE_CUDA \
-DGGML_BUILD=1 \
-DGGML_BACKEND_BUILD=1 \
-DGGML_SHARED=1 \
-DGGML_BACKEND_SHARED=1 \
-DGGML_CUDA_DMMV_X=32 \
-DGGML_CUDA_MMV_Y=1 \
-DGGML_SCHED_MAX_COPIES=4 \
-DGGML_USE_HIPBLAS \
-DGGML_USE_HIP \
-DGGML_USE_LLAMAFILE \
-DHIP_FAST_MATH \
-D__HIP_PLATFORM_AMD__=1 \