Skip to content

[Auto-parallel] Adapt ernie pp for auto_dy #73615

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 5 commits into from
Jul 23, 2025
Merged

[Auto-parallel] Adapt ernie pp for auto_dy #73615

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
8af3fe0
Update microbatch.py
Xing-lil Jun 25, 2025
ce52dcf
Update schedules.py
Xing-lil Jun 25, 2025
b9d4bfb
Update microbatch.py
Xing-lil Jun 25, 2025
0aedbe4
Merge branch 'PaddlePaddle:develop' into ernie_pp
Xing-lil Jun 26, 2025
236416e
Update schedules.py
Xing-lil Jul 21, 2025
File filter

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions python/paddle/distributed/auto_parallel/pipelining/microbatch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -142,13 +142,15 @@ def _split_args_helper(
for idx in range(num_chunks):
chunk_args = {}
for key, arg in shared_args_dict_flat.items():
last_arg = None if not arg else arg[0][idx]
arg_of_curr_chunk = (
[v[idx] for v in arg] if len(arg) > 1 else arg[0][idx]
[v[idx] for v in arg] if len(arg) > 1 else last_arg
)
chunk_args[key] = arg_of_curr_chunk

# flatten chunk_args first, and then pack chunk_args as the origin args_dict
chunk_args = pack_sequence_as(args_dict, flatten(chunk_args))
flatten_chunk_args = [x for x in flatten(chunk_args) if x is not None]
chunk_args = pack_sequence_as(args_dict, flatten_chunk_args)
args_split.append(chunk_args)
return args_split

Expand Down
Loading