Skip to content

[HOTFIX MGPU unit test] for test_cast_master_weights_to_fp8.py #491

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

Merged
sudhu2k merged 2 commits into from
Mar 18, 2026
Merged

[HOTFIX MGPU unit test] for test_cast_master_weights_to_fp8.py #491

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
882f2f7
Refactor test_cast_master_weights_to_fp8 to use executable directly w…
sudhu2k Mar 17, 2026
398ba85
Enhance test_cast_master_weights_to_fp8 for AMDGPU portability and up…
sudhu2k Mar 18, 2026
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
6 changes: 5 additions & 1 deletion tests/pytorch/distributed/test_cast_master_weights_to_fp8.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,5 @@
# This file was modified for portability to AMDGPU.
# Copyright (c) 2026, Advanced Micro Devices, Inc. All rights reserved.
# Copyright (c) 2022-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# See LICENSE for license information.
Expand Down Expand Up @@ -717,7 +719,9 @@ def run_parallel_tests() -> None:
@pytest.mark.parametrize("world_size", [2])
def test_cast_master_weights_to_fp8(world_size: int) -> None:
"""Launch parallel job that runs parallel tests"""
python_exe = pathlib.Path(sys.executable).resolve()
Copy link
Collaborator

@ipanfilo ipanfilo Mar 18, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it cherry-pick from upstream or our change? In the latter case please update copyright and would be nice to tag comment with ROCm or AMD so it is not accidentally removed on further IFUs

# ROCm: Use executable as-is; do not resolve() or a venv symlink may point to system
# Python which does not have torch/site-packages.
python_exe = pathlib.Path(sys.executable)
current_file = pathlib.Path(__file__).resolve()
command = [
python_exe,
Expand Down