Skip to content
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

Jump to bottom

[Bug Fix] fix allreduce tensor dtype #7876

Merged
merged 2 commits into from
Jan 25, 2024
Merged

[Bug Fix] fix allreduce tensor dtype #7876

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a5879d3
[Bug Fix] fix allreduce tensor dtype
BeingGod Jan 22, 2024
0e751e8
update int8 to int32
BeingGod Jan 22, 2024
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 paddlenlp/trainer/trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -626,7 +626,7 @@
# The resume_from_checkpoint could be None in some machine node.
# Here we reset None to temp directory.
if args.world_size > 1:
is_resume_from_checkpoint = paddle.to_tensor([resume_from_checkpoint is not None])
is_resume_from_checkpoint = paddle.to_tensor([resume_from_checkpoint is not None], dtype="int8")

Check warning on line 629 in paddlenlp/trainer/trainer.py

View check run for this annotation

Codecov / codecov/patch

paddlenlp/trainer/trainer.py#L629 

Added line #L629 was not covered by tests
BeingGod marked this conversation as resolved.
Show resolved Hide resolved
paddle.distributed.all_reduce(is_resume_from_checkpoint)
is_resume_from_checkpoint = is_resume_from_checkpoint.item()
if is_resume_from_checkpoint > 0 and is_resume_from_checkpoint < paddle.distributed.get_world_size():
Expand Down