Device selection with pytest. #1713
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
You are owner of some files modified in this pull request. |
boring-cyborg
bot
added
test-cases
gabrieldemarmiesse
requested review from
AakashKumarNain,
guillaumekln,
qlzh727,
Squadrick and
WindQAQ
as code owners
seanpmorgan
requested changes
May 1, 2020
There was a problem hiding this comment.
Looks very good thanks for continuing to build out our pytest suite! I tested locally and got expected results. Small doc change requested.
I also had trouble finding where the op logging was coming from. For example at the end of a test run it'll ouput:
transform/strided_slice_2/stack_1: (Const): /job:localhost/replica:0/task:0/device:GPU:0
transform/strided_slice_2/stack_2: (Const): /job:localhost/replica:0/task:0/device:GPU:0
transform/assert_rank/rank: (Const): /job:localhost/replica:0/task:0/device:GPU:0
transform/assert_rank/Shape: (Const): /job:localhost/replica:0/task:0/device:GPU
...
This seems great for debugging, but it is a bit unwieldy to look through if your testing something other than device placement.
Co-authored-by: Sean Morgan <[email protected]>
…s into add_pytest_mark
|
Yeah, I totally forgot to remove the logging of the devices. It's done now. |
ashutosh1919
pushed a commit
to ashutosh1919/addons
that referenced
this pull request
Jul 12, 2020
* The device to execute each test is set in stone. * Update CONTRIBUTING.md * Removed device placement logging and added comment.
jrruijli
pushed a commit
to jrruijli/addons
that referenced
this pull request
Dec 23, 2020
* The device to execute each test is set in stone. * Update CONTRIBUTING.md * Removed device placement logging and added comment.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes part of #1682
Note that I left out the idea of using many gpus for now as it's not necessary. One is more than enough.
This focuses on enabling multiprocessing with the gpu available.
This also have another benefit: it doesn't matter if your system has a gpu or not anymore, the same tests will get exectuted on the same devices.
We can now use many workers, even when only one GPU is available to run the tests.
It's compatible with calling manually
with tf.device("GPU:0"), but I believe it's cleaner to use this built-in solution so we should replace the leftovers later on.Also in the TODO, instead of using one virtual device per process, we can make multiple ones, and then do something like
@pytest.mark.with_device(["mirror_strategy"]).