Skip to content

[model] optimize padding mask reduction memory #167

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
Zhijun-Xu wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[model] optimize padding mask reduction memory #167

Show file tree
Hide file tree
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
2 changes: 1 addition & 1 deletion src/mcore_bridge/model/gpt_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -300,7 +300,7 @@ def forward(
full_attention_mask = full_attention_mask['full_attention']
if mcore_016 and full_attention_mask is not None:
assert packed_seq_params is None
padding_mask = ~((~full_attention_mask).sum(dim=(1, 2)) > 0)
padding_mask = full_attention_mask.all(dim=(1, 2))
if self.config.context_parallel_size > 1:
padding_mask = split_cp_inputs(padding_mask, None, 1)
tp_size = self.config.tensor_model_parallel_size
Expand Down
2 changes: 1 addition & 1 deletion src/mcore_bridge/model/hybrid_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,7 +59,7 @@ def _get_padding_mask(self, attention_mask) -> Optional[torch.Tensor]:
attention_mask = attention_mask['full_attention']
if attention_mask is None:
return None
padding_mask = ~((~attention_mask).sum(dim=(1, 2)) > 0)
padding_mask = attention_mask.all(dim=(1, 2))
if self.config.context_parallel_size > 1:
padding_mask = split_cp_inputs(padding_mask, None, 1)
tp_size = self.config.tensor_model_parallel_size
Expand Down