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

Refactor Error Handling for CUDA

Browse source 
Add guidance for setting CUDA_DOCKER_ARCH to match GPU compute capability for CUDA versions < 11.7. Include link to NVIDIA's CUDA GPUs documentation for compute capability reference.
This commit is contained in:
Nikolas 2024-04-10 03:17:59 +02:00 • committed by GitHub
parent ba5e134e07
commit 377a52427c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
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

@ -646,7 +646,7 @@ CUDA_VERSION := $(shell $(NVCC) --version | grep -oP 'release (\K[0-9]+\.[0-9])'
ifeq ($(shell awk -v "v=$(CUDA_VERSION)" 'BEGIN { print (v < 11.7) }'),1)
ifndef CUDA_DOCKER_ARCH
ifndef CUDA_POWER_ARCH
$(error I ERROR: For CUDA versions < 11.7 a target CUDA architecture must be explicitly provided via CUDA_DOCKER_ARCH)
$(error I ERROR: To address this error for CUDA versions earlier than 11.7, you must specify your GPU's compute capability. Execute export CUDA_DOCKER_ARCH=compute_XX, replacing XX with your GPU's compute capability number. For assistance in finding your GPU's compute capability, visit https://developer.nvidia.com/cuda-gpus.)
endif # CUDA_POWER_ARCH
endif # CUDA_DOCKER_ARCH
endif # eq ($(shell echo "$(CUDA_VERSION) < 11.7" | bc),1)