Skip to content

test_fn_fwgrad_bwgrad_[trapezoid|trapz]_cuda_complex128 causes CUDA memory exception #72189

New issue
New issue
Open
Open
test_fn_fwgrad_bwgrad_[trapezoid|trapz]_cuda_complex128 causes CUDA memory exception#72189
Labels
ZeroTensormodule: ciRelated to continuous integrationmodule: crashProblem manifests as a hard crash, as opposed to a RuntimeErrormodule: cudaRelated to torch.cuda, and CUDA support in generaltriagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate module
@malfet

Description

@malfet
malfet
opened on Feb 2, 2022

🐛 Describe the bug

Discovered while running #72016 that adds explicit error even if CUDA memory exception caused early termination
See this runlog for trapezoid failure: https://github.com/pytorch/pytorch/runs/5040778523?check_suite_focus=true
And this one for trapz: https://github.com/pytorch/pytorch/runs/5041795069?check_suite_focus=true

Versions

N/A

cc @ngimel @seemethere @malfet @pytorch/pytorch-dev-infra

Metadata

Metadata

Assignees

No one assigned

    Labels

    ZeroTensormodule: ciRelated to continuous integrationmodule: crashProblem manifests as a hard crash, as opposed to a RuntimeErrormodule: cudaRelated to torch.cuda, and CUDA support in generaltriagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions