Fix attention fusion in conformer encoder#23711
Merged
kunal-vaishnavi merged 1 commit intomicrosoft:mainfrom Feb 16, 2025
Merged
Conversation
hanbitmyths
approved these changes
Feb 15, 2025
jingyanwangms
added a commit
that referenced
this pull request
Feb 18, 2025
This reverts commit 47a0077.
guschmue
pushed a commit
that referenced
this pull request
Mar 6, 2025
### Description This PR updates the attention fusion for conformer-encoder models. It is a follow-up to [this PR](#23528). ### Motivation and Context Subsequent modeling code updates have changed (and will continue to change) the graph fusions. However, the three ending attention mask nodes (`Cast --> Unsqueeze --> Equal`) will remain. Thus, the attention fusion should work regardless of any future modeling code changes when handling the attention mask.
ashrit-ms
pushed a commit
that referenced
this pull request
Mar 17, 2025
### Description This PR updates the attention fusion for conformer-encoder models. It is a follow-up to [this PR](#23528). ### Motivation and Context Subsequent modeling code updates have changed (and will continue to change) the graph fusions. However, the three ending attention mask nodes (`Cast --> Unsqueeze --> Equal`) will remain. Thus, the attention fusion should work regardless of any future modeling code changes when handling the attention mask.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
This PR updates the attention fusion for conformer-encoder models. It is a follow-up to this PR.
Motivation and Context
Subsequent modeling code updates have changed (and will continue to change) the graph fusions. However, the three ending attention mask nodes (
Cast --> Unsqueeze --> Equal) will remain. Thus, the attention fusion should work regardless of any future modeling code changes when handling the attention mask.