Skip to content

[BugFix] Fix attention mask for multimodal models #7841

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
TBD1 wants to merge 1 commit into
base: develop
Choose a base branch
from
+9 −3
Open

[BugFix] Fix attention mask for multimodal models #7841

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
12 changes: 9 additions & 3 deletions fastdeploy/spec_decode/mtp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -523,10 +523,16 @@ def insert_tasks_v1(
)
if self.use_attn_mask_offset:
inputs = request.multimodal_inputs
self.model_inputs["attn_mask_offsets_full"][idx][0 : prefill_end_index - prefill_start_index] = (
paddle.to_tensor(
inputs["attention_mask_offset"][prefill_start_index:prefill_end_index], dtype="int32"
attn_offset_len = prefill_end_index - prefill_start_index
if inputs.get("attention_mask_offset", None) is None:
attention_mask_offset_slice = np.arange(prefill_start_index, prefill_end_index, dtype=np.int32)
else:
attention_mask_offset_slice = np.asarray(
inputs["attention_mask_offset"][prefill_start_index:prefill_end_index],
dtype=np.int32,
)
self.model_inputs["attn_mask_offsets_full"][idx][0:attn_offset_len] = paddle.to_tensor(
attention_mask_offset_slice, dtype="int32"
)
# GPU don't need it anymore
# NOTE: XPU backend needs decoder attention mask offset; GPU backend does not use it
Expand Down
Loading