TF: rework XLA generate tests #16866
Merged
Conversation
This file contains 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
|
The documentation is not available anymore as the PR was closed or merged. |
Comment on lines
+483
to
+486
| expected_output_string = [ | ||
| "Heute ist ein schöner Tag.", | ||
| "Ich habe vier Katzen, drei Hunde, zwei Vögel und ein Pferd.", | ||
| ] |
There was a problem hiding this comment.
Are we guaranteed that the logits are stable enough that we'll always sample this exact output? A flaky test can be really annoying!
There was a problem hiding this comment.
We haven't had problems with similar tests, so I'm assuming we won't have problems :D
Rocketknight1
approved these changes
Apr 21, 2022
patrickvonplaten
approved these changes
Apr 22, 2022
|
@patrickvonplaten reintroduced the fast tests, will merge as soon as CI gets to green |
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Outdated suggestions cannot be applied.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.
What does this PR do?
In the light of recent findings (#16838), this PR reworks existing XLA generate tests. The following key changes were made:
@unittest.skipIfon XLA generate tests, to skip when no GPU is present;sampletests -- due to the minor numerical differences that arise when we use XLA (and that we can't control), the sampling step will gather different samples even when we use the same seed. The only thing we can properly test is whether a) we can seed them and b) the results are sensible;