Merge key_padding_mask into attn_mask_rel_pos in WavLM #3265
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
|
@nateanl has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@nateanl has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
Hey @nateanl. |
nateanl
added a commit
to nateanl/audio
that referenced
this pull request
Apr 12, 2023
Summary: When `key_padding_mask` is not `None`, it needs to be combined with `attn_mask_rel_pos` as one mask for `scaled_dot_product_attention` function. Pull Request resolved: pytorch#3265 Reviewed By: hwangjeff Differential Revision: D44901093 Pulled By: nateanl fbshipit-source-id: 73ca7af48faf7f4eb36b35b603187a11e5582c70
nateanl
added a commit
to nateanl/audio
that referenced
this pull request
Apr 12, 2023
Summary: When `key_padding_mask` is not `None`, it needs to be combined with `attn_mask_rel_pos` as one mask for `scaled_dot_product_attention` function. Pull Request resolved: pytorch#3265 Reviewed By: hwangjeff Differential Revision: D44901093 Pulled By: nateanl fbshipit-source-id: 73ca7af48faf7f4eb36b35b603187a11e5582c70
nateanl
added a commit
that referenced
this pull request
Apr 12, 2023
, #3265) (#3264) * Use scaled_dot_product_attention in WavLM attention (#3252) Summary: Fix #3219. `torch.nn.MultiheadAttention` will throw an error if `torch.no_grad()` and mask are both given. The pull request fixes it by replacing the forward method with `torch.nn.functional.scaled_dot_product_attention`. Pull Request resolved: #3252 Reviewed By: mthrok Differential Revision: D44798634 Pulled By: nateanl fbshipit-source-id: abfa7fb84b7bd71848a92ab26da5a5f0f095c665 * Merge key_padding_mask into attn_mask_rel_pos in WavLM (#3265) Summary: When `key_padding_mask` is not `None`, it needs to be combined with `attn_mask_rel_pos` as one mask for `scaled_dot_product_attention` function. Pull Request resolved: #3265 Reviewed By: hwangjeff Differential Revision: D44901093 Pulled By: nateanl fbshipit-source-id: 73ca7af48faf7f4eb36b35b603187a11e5582c70
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.
When
key_padding_maskis notNone, it needs to be combined withattn_mask_rel_posas one mask forscaled_dot_product_attentionfunction.