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

[Distributed]Add tensorfusion bug report #59762

Merged
merged 1 commit into from
Dec 7, 2023
Merged

[Distributed]Add tensorfusion bug report #59762

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
5 changes: 3 additions & 2 deletions python/paddle/distributed/fleet/utils/tensor_fusion_helper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -372,8 +372,9 @@ def add_grad(self, param, use_comm=True):
raise ValueError(
"The address of the grad/main_grad of the param has been changed during training, "
"which is not allowed for dp/sharding overlap with pp. "
"This may be caused by some non-inplace operations on the grad/main_grad. "
"Please use the inplace version of the operations or disable the overlapping."
"This may be caused by some non-inplace operations on the grad/main_grad. Here are some examples: "
"1. The grad/main_grad of the param is changed by other operations, such as: clear_grad, "
"2. Using non-inplace operations on the grad/main_grad, such as: add, sub, mul, div, etc. "
)

self._params_step_dict[param.name] += 1
Expand Down