[fx] Add common node in model linearize #1542
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
Merge ColossalAI
Daily merge
super-dainiu
reviewed
Sep 5, 2022
…her30/ColossalAI into feature/add_common_node_to_linearize
super-dainiu
approved these changes
Sep 5, 2022
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Outdated suggestions cannot be applied.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's New
We were inspired by the common node idea in merak, and modify the naive process of defining a common node in merak's source code. Now we are able to address the attention mask problem in hugginface BERT model, separating each attention block and MLP in the encoder.
Moreover, I also fixed the problem that our solver might repeatedly count the memory for combination like "BN-RELU", as the current linearize strategy will generate more isolated in-place RELU node in the linearized node list.