Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make padding_mask Optional in SequenceBatch. #106

Merged
merged 2 commits into from
Oct 17, 2023
Merged

Make padding_mask Optional in SequenceBatch. #106

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
115c349
Make padding_mask Optional in SequenceBatch.
kauterry Oct 17, 2023
5ca38ae
Make compute_num_tokens() in SequenceBatch to account for the corner …
kauterry Oct 17, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion src/fairseq2/models/sequence.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,7 +37,7 @@ class SequenceBatch:
size, :math:`S` is the sequence length, and :math:`*` is any number of
sequence-specific dimensions including none."""

padding_mask: PaddingMask
padding_mask: Optional[PaddingMask]
"""The padding mask of ``seqs``. *Shape:* :math:`(N,S)`, where :math:`N` is
the batch size and :math:`S` is the sequence length."""

Expand All @@ -51,6 +51,9 @@ def batch_size(self) -> int:

def compute_num_tokens(self) -> Tensor:
"""Compute the number of tokens in this batch."""
if self.padding_mask is None:
return torch.full((), self.seqs.numel(), device=self.seqs.device)

return self.padding_mask.seq_lens.sum()


Expand Down