Skip to content

Fix return type annotation for _get_sentence_embedding to match dict return #68

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

Merged
meilame-tayebjee merged 4 commits into from
Jan 27, 2026
Merged

Fix return type annotation for _get_sentence_embedding to match dict return #68

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
fe35fc2
Initial plan
Copilot Jan 27, 2026
63cb646
Update _get_sentence_embedding return type annotation and docstring
Copilot Jan 27, 2026
e9635ea
Fix early returns to match dictionary return type
Copilot Jan 27, 2026
fb4b9fa
Merge branch '24-add-cross-attention-labels-text' into copilot/sub-pr…
meilame-tayebjee Jan 27, 2026
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
29 changes: 19 additions & 10 deletions torchTextClassifiers/model/components/text_embedder.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
import math
from dataclasses import dataclass
from typing import Optional
from typing import Dict, Optional

import torch
import torch.nn as nn
Expand Down Expand Up @@ -129,7 +129,7 @@ def forward(
input_ids: torch.Tensor,
attention_mask: torch.Tensor,
return_label_attention_matrix: bool = False,
) -> dict[str, Optional[torch.Tensor]]:
) -> Dict[str, Optional[torch.Tensor]]:
"""Converts input token IDs to their corresponding embeddings.

Args:
Expand Down Expand Up @@ -200,15 +200,18 @@ def _get_sentence_embedding(
token_embeddings: torch.Tensor,
attention_mask: torch.Tensor,
return_label_attention_matrix: bool = False,
) -> torch.Tensor:
) -> Dict[str, Optional[torch.Tensor]]:
"""
Compute sentence embedding from embedded tokens - "remove" second dimension.

Args (output from dataset collate_fn):
token_embeddings (torch.Tensor[Long]), shape (batch_size, seq_len, embedding_dim): Tokenized + padded text
attention_mask (torch.Tensor[Long]), shape (batch_size, seq_len): Attention mask indicating non-pad tokens
return_label_attention_matrix (bool): Whether to compute and return the label attention matrix
Returns:
torch.Tensor: Sentence embeddings, shape (batch_size, embedding_dim)
Dict[str, Optional[torch.Tensor]]: A dictionary containing:
- 'sentence_embedding': Sentence embeddings, shape (batch_size, embedding_dim) or (batch_size, n_labels, embedding_dim) if label attention is enabled
- 'label_attention_matrix': Attention matrix if label attention is enabled and return_label_attention_matrix is True, otherwise None
"""

# average over non-pad token embeddings
Expand All @@ -219,14 +222,20 @@ def _get_sentence_embedding(
if self.attention_config is not None:
if self.attention_config.aggregation_method is not None: # default is "mean"
if self.attention_config.aggregation_method == "first":
return token_embeddings[:, 0, :]
return {
"sentence_embedding": token_embeddings[:, 0, :],
"label_attention_matrix": None,
}
elif self.attention_config.aggregation_method == "last":
lengths = attention_mask.sum(dim=1).clamp(min=1) # last non-pad token index + 1
return token_embeddings[
torch.arange(token_embeddings.size(0)),
lengths - 1,
:,
]
return {
"sentence_embedding": token_embeddings[
torch.arange(token_embeddings.size(0)),
lengths - 1,
:,
],
"label_attention_matrix": None,
}
else:
if self.attention_config.aggregation_method != "mean":
raise ValueError(
Expand Down