Skip to content

ZeRO3: Gradient norm allreduce for DP #1021

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
tjruwase merged 2 commits into from
Apr 30, 2021
Merged

ZeRO3: Gradient norm allreduce for DP #1021

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a9879b8
Handle Norm allreduce when no mp
tjruwase Apr 29, 2021
3e11f4c
Merge branch 'master' into olruwase/zero3_dp_norm_allreduce
jeffra Apr 29, 2021
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 deepspeed/runtime/zero/stage3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2405,7 +2405,7 @@ def _model_parallel_all_reduce(self, tensor, op):
""" Perform all reduce within model parallel group, if any.
"""
if self.model_parallel_group is None:
torch.distributed.all_reduce(tensor=tensor, op=op)
pass
else:
torch.distributed.all_reduce(tensor=tensor,
op=op,
Expand Down