Skip to content

fix: restore compatibility with transformers >= 4.49 #458

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

Open
phreed wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: restore compatibility with transformers >= 4.49 #458

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
13 changes: 11 additions & 2 deletions groundingdino/models/GroundingDINO/bertwarper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,7 +26,13 @@ def __init__(self, bert_model):

self.get_extended_attention_mask = bert_model.get_extended_attention_mask
self.invert_attention_mask = bert_model.invert_attention_mask
self.get_head_mask = bert_model.get_head_mask
# Check if the attribute exists; if not, set it to a dummy function or None
if hasattr(bert_model, "get_head_mask"):
self.get_head_mask = bert_model.get_head_mask
else:
# Newer transformers versions might not have this,
# but GroundingDINO's forward pass might call it.
self.get_head_mask = lambda x, y: None

def forward(
self,
Expand Down Expand Up @@ -106,8 +112,11 @@ def forward(

# We can provide a self-attention mask of dimensions [batch_size, from_seq_length, to_seq_length]
# ourselves in which case we just need to make it broadcastable to all heads.
# Pass the dtype of the model weights instead of the device object
model_dtype = next(self.parameters()).dtype

extended_attention_mask: torch.Tensor = self.get_extended_attention_mask(
attention_mask, input_shape, device
attention_mask, input_shape, model_dtype
)

# If a 2D or 3D attention mask is provided for the cross-attention
Expand Down