Skip to content

Intermittent failure of sebotnet33ts_256 on CI #94066

New issue
New issue
Closed
Closed
Intermittent failure of sebotnet33ts_256 on CI#94066
Assignees
desertfire
Labels
module: flaky-testsProblem is a flaky test in CImodule: testingIssues related to the torch.testing module (not tests)oncall: pt2triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate module
@desertfire

Description

@desertfire
desertfire
opened on Feb 3, 2023

It has existed for a while and happens more frequently lately, https://hud.pytorch.org/failure/sebotnet33ts_256%2Cfail_accuracy

Repro:

python benchmarks/dynamo/timm_models.py --ci --training --accuracy --device cuda --inductor --amp --only sebotnet33ts_256

cc @ezyang @soumith @msaroufim @wconstab @ngimel @bdhirsh

Metadata

Metadata

Assignees

Labels

module: flaky-testsProblem is a flaky test in CImodule: testingIssues related to the torch.testing module (not tests)oncall: pt2triagedThis 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