Skip to content

Make sharding_first by default #2822

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
lugimzzz merged 1 commit into from
Nov 3, 2025
Merged

Make sharding_first by default #2822

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
4 changes: 2 additions & 2 deletions paddleformers/trainer/training_args.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -783,7 +783,7 @@ class TrainingArguments:
"Following options are supported:\n"
"- pp_first. the topo order is dp, pp, sharding, mp \n"
"- sharding_first. the topo order is dp, sharding, pp, mp \n"
"Default is None, for pp_first"
"Default is None, for sharding_first"
)
},
)
Expand Down Expand Up @@ -2132,7 +2132,7 @@ def _post_init_parallel_degree(self):
self.expert_tensor_parallel_degree = -1

if self.hybrid_parallel_topo_order is None:
self.hybrid_parallel_topo_order = "pp_first"
self.hybrid_parallel_topo_order = "sharding_first"
assert self.hybrid_parallel_topo_order in ["pp_first", "sharding_first"]

if self.use_hybrid_parallel and self.enable_auto_parallel:
Expand Down