Skip to content

expose TORCH_CUDA_ARCH_LIST #147

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
ErwannMillon wants to merge 2 commits into
base: main
Choose a base branch
from
Open

expose TORCH_CUDA_ARCH_LIST #147

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
8131369
expose TORCH_CUDA_ARCH_LIST
ErwannMillon Mar 24, 2025
adda197
override automatically detected cuda architectures if TORCH_CUDA_ARCH…
ErwannMillon Mar 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions setup.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,7 @@
SUPPORTED_ARCHS = {"8.0", "8.6", "8.9", "9.0", "12.0"}

# Compiler flags.
TORCH_CUDA_ARCH_LIST = os.environ.get("TORCH_CUDA_ARCH_LIST", "")
CXX_FLAGS = ["-g", "-O3", "-fopenmp", "-lgomp", "-std=c++17", "-DENABLE_BF16"]
NVCC_FLAGS = [
"-O3",
Expand Down Expand Up @@ -76,6 +77,9 @@ def get_nvcc_cuda_version(cuda_dir: str) -> Version:
continue
compute_capabilities.add(f"{major}.{minor}")

if TORCH_CUDA_ARCH_LIST:
compute_capabilities = set(TORCH_CUDA_ARCH_LIST.split(","))

nvcc_cuda_version = get_nvcc_cuda_version(CUDA_HOME)
if not compute_capabilities:
raise RuntimeError("No GPUs found. Please specify the target GPU architectures or build on a machine with GPUs.")
Expand Down