Skip to content
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

Jump to bottom

[Fix] Disable CUDA multiarch fatbin by default #716

Merged
merged 1 commit into from Aug 9, 2023
Merged

[Fix] Disable CUDA multiarch fatbin by default #716

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
4 changes: 2 additions & 2 deletions mlc_llm/utils.py
View file
Open in desktop
Expand Up @@ -366,7 +366,7 @@ def parse_target(args: argparse.Namespace) -> None:
)
args.target = target
args.target_kind = args.target.kind.default_keys[0]
elif args.target == "cuda":
elif args.target == "cuda" or args.target == "cuda-multiarch":
target = _detect_local_cuda()
if target is None:
raise ValueError("Cannot detect local CUDA GPU target!")
Expand Down Expand Up @@ -513,7 +513,7 @@ def compile_metal(src, target):
args.target = tvm.target.Target(args.target, host="llvm")
args.target_kind = args.target.kind.default_keys[0]

if args.target_kind == "cuda":
if args.target_kind == "cuda-multiarch":
from tvm.contrib import nvcc

assert args.target.arch[3:] != ""
Expand Down