Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Torch backwards compatibility fixes #610

Merged
merged 8 commits into from
Mar 15, 2022
Merged

Torch backwards compatibility fixes #610

merged 8 commits into from
Mar 15, 2022

Conversation

danieldk
Copy link
Contributor

  • Fix compatibility with Torch without torch.cuda.amp.common
  • Disable PyTorch-based activation tests pre-PyTorch 1.9.0

@danieldk danieldk mentioned this pull request Mar 14, 2022
Closed
@svlandeg svlandeg added the interop / pytorch PyTorch interoperability label Mar 14, 2022
@danieldk
Copy link
Contributor Author

@explosion-bot please test_gpu

@explosion-bot
Copy link
Collaborator

explosion-bot commented Mar 14, 2022 •
edited
Loading

🪁 Successfully triggered build on Buildkite

URL: https://buildkite.com/explosion-ai/thinc-gpu-test-suite/builds/32

This was referenced Mar 14, 2022
Merged
Closed
@danieldk danieldk marked this pull request as draft March 15, 2022 09:40
@danieldk
Check that torch>=1.9.0 for mixed-precision training
53089c8 
Torch versions prior to 1.9.0 do not have the functionality that we need
for mixed-precision training and gradient scaling.
@danieldk danieldk marked this pull request as ready for review March 15, 2022 10:23
@danieldk danieldk merged commit 25ac7f9 into explosion:master Mar 15, 2022
@danieldk danieldk deleted the torch-backward-compat branch March 15, 2022 11:49
danieldk added a commit that referenced this pull request Mar 16, 2022
@danieldk
Torch backwards compatibility fixes (#610)
f50905f 
* Fix compatibility with Torch without torch.cuda.amp.common

* Disable PyTorch-based activation tests pre-PyTorch 1.9.0

* Don't use gradient scaling unconditionally in PyTorch wrapper test

* Disable gradient scaling tests on older PyTorch versions

* Set minimum required PyTorch version to 1.6.0

* Check that torch>=1.9.0 for mixed-precision training

Torch versions prior to 1.9.0 do not have the functionality that we need
for mixed-precision training and gradient scaling.

* Refine exception message for mixed-precision training
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
interop / pytorch PyTorch interoperability
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@danieldk @explosion-bot @svlandeg