Skip to content

Conversation

elalish
Copy link
Contributor

@elalish elalish commented Sep 3, 2021

We've had a lot of weird flakiness in our fidelity tests that we didn't understand, so we just turned the flake into a warning so we could ignore it. However, the fidelity tests actually catch quite a few nasty regressions, so I'd rather they were all functional. I found a few sketchy things that I've fixed here. I was only able to get a consistent repro on a single model locally, and this fixed that, but we'll see if the CI agrees.

And it looks like it does! Seems the catch for the poster issue must have been somehow wrong, but there's no poster problem now and once that check was removed everything seems to be working well.

@elalish elalish self-assigned this Sep 3, 2021
@elalish elalish requested a review from sun765 September 3, 2021 22:45
@elalish elalish merged commit 7563734 into master Sep 3, 2021
@elalish elalish deleted the fidelityTest branch September 3, 2021 23:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant