Skip to content

Conversation

@frankyoujian
Copy link
Contributor

Issue Description:
When I want to dump features at the middle layer of encoder, i.e. zipformer, I will call extract_features like fairseq dump_hubert_feature . However, current encoder forward calling does not support None padding_mask.

Fix:
Handle None padding_mask when invoking encoder forward calling

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant