vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update Makefile for misc amd gpu targetting (#407)

Browse source 
adds the hipBlas gpu_target $(shell $(ROCM_PATH)/llvm/bin/amdgpu-arch)
back to the gpu_target line, possibly allowing misc gpu arch's like gfx1031 or gfx1032 etc to be built
This commit is contained in:
YellowRoseCx 2023-08-29 20:54:15 -05:00 • committed by GitHub
parent 380fa0f0ca
commit d7bdfbdd78
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
Makefile
View file

@ -197,7 +197,7 @@ ifdef LLAMA_HIPBLAS
ROCM_PATH ?= /opt/rocm
CC := $(ROCM_PATH)/llvm/bin/clang
CXX := $(ROCM_PATH)/llvm/bin/clang++
GPU_TARGETS ?= gfx803 gfx900 gfx906 gfx908 gfx90a gfx1030 gfx1100
GPU_TARGETS ?= gfx803 gfx900 gfx906 gfx908 gfx90a gfx1030 gfx1100 $(shell $(ROCM_PATH)/llvm/bin/amdgpu-arch)
LLAMA_CUDA_DMMV_X ?= 128
LLAMA_CUDA_MMV_Y ?= 2
LLAMA_CUDA_KQUANTS_ITER ?= 1