Skip to content

[PEFT]fix lora model tp when existing other trainable module #7781

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
merged 1 commit into from
Jan 8, 2024
Merged

[PEFT]fix lora model tp when existing other trainable module #7781

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
3 changes: 1 addition & 2 deletions paddlenlp/peft/lora/lora_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -195,8 +195,7 @@ def _merge_trainable_tensor_parallel(self, trainable_state_dict):
if key in trainable_name_action_mappings:
ret = distributed_gather(tensor, group=mp_group, offload=True)
action = trainable_name_action_mappings[key]
is_collumn = self.lora_split_mapping[key]
if "_scale" in key and not is_collumn and is_dst:
if key in self.lora_split_mapping and not self.lora_split_mapping[key] and "_scale" in key and is_dst:
ret = paddle.to_tensor(ret)
tensor = paddle.max(ret, axis=0)
else:
Expand Down