Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Add pytest marker for GPU tests

See original GitHub issue

It came up in the monthly community meeting that we should add a gpu pytest marker to certain test modules / individual tests which run on GPUs. This will allow us to select only GPU-related tests when running the test suite on machines with GPUs in CI (xref https://github.com/dask/distributed/pull/5011)

cc @charlesbluca @quasiben @aktech

Issue Analytics

State:
Created 2 years ago
Reactions:1
Comments:8 (7 by maintainers)

Top GitHub Comments

2reactions

jrbourbeaucommented, Jul 1, 2021

Great! If I were doing this I would probably do something like:

Add a description for a new gpu marker here

https://github.com/dask/distributed/blob/8b79bd95d0c5a681f6d0543eb20d645dc1182ec1/setup.cfg#L49-L52

this will inform pytest about the new marker

Mark all of the test modules here (except for test_nanny.py) with a module-level pytest.mark.gpu marker
Go into test_nanny.py and decorate individual tests that run on GPUs with @pytest.mark.gpu

1reaction

quasibencommented, Jul 6, 2021

We added test_nanny because of a UCX issue. Here is the PR for the work https://github.com/dask/distributed/pull/3747 – I think if it’s not too much overhead testing the entire file would be good

Top Results From Across the Web

Testing — transformers 3.4.0 documentation - Hugging Face

pytest -sugar is a plugin that improves the look-n-feel, adds a progressbar, and show tests that fail and the assert instantly.

How to mark test functions with attributes - Pytest

By using the pytest.mark helper you can easily set metadata on your test functions. You can find the full list of builtin markers...

Enable multiprocessing when testing with GPU and support ...

To do that we'll use pytest markers: @pytest.mark..... By default, if no marker is found, the test will run on CPU: with device(" ......

pytest-manual-marker - PyPI

Pytest marker for marking manual tests. Adds options for collecting manual, automated (default) or both. Adds a different test outcome for manaual tests....

run dependant tests with pytest - Stack Overflow

If you filter by markers, this is not possible, AFAIK. Can you elaborate on why you can't add the marker to test_a ?...