Flaky test - tests/models/mobilenet_v1/test_modeling_mobilenet_v1.py::MobileNetV1ModelTest::test_batching_equivalence

[amyeroberts]

System Info

tests/models/mobilenet_v1/test_modeling_mobilenet_v1.py::MobileNetV1ModelTest::test_batching_equivalence periodically fails causing red on CI runs

Example CI run: https://app.circleci.com/pipelines/github/huggingface/transformers/94887/workflows/6766b7cd-c38c-4334-ab25-2afcb8e0d842/jobs/1246458

Who can help?

@zucchini-nlp as you added the batching equivalence tests :)

Information

• [ ] The official example scripts
• [ ] My own modified scripts

Tasks

• [ ] An officially supported task in the examples folder (such as GLUE/SQuAD, ...)
• [ ] My own task or dataset (give details below)

Reproduction

Run enough times to hit an error

Expected behavior

Doesn't occasionally fail

[zucchini-nlp]

Yeah, probably it was flaky at the time of adding because the MobileNetV2 is also flaky as marked #29516

As discussed, I am not CV expert but maybe I can give it another look later next week