Better dispatch for submodules #392
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. |
LysandreJik
approved these changes
May 24, 2022
|
Not yet, will have a look before merging (though I don't expect much difference since there is not that many submodules). |
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Outdated suggestions cannot be applied.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.
While testing the
dispatch_modelfeatures in this PR I hit some edge case for model parallelism: when a submodule of a given model is never actually called in the forward pass (like thehof the GPT-2 model which is a module list which is iterated on in the forward but not directly called) the inputs aren't necessarily moved to the right device.This PR adds a hook in all submodules to just move those inputs to the device, which fix this issue.