Skip to content

Restore missing linear-cross-entropy option accidentally removed from arguments.py #3266

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
yaox12 merged 2 commits into from
Feb 6, 2026
Merged

Restore missing linear-cross-entropy option accidentally removed from arguments.py #3266

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a0c4904
Add missing linear-cross-entropy option
shjwudp Feb 5, 2026
7ff5ba0
Merge branch 'dev' into entry_flag_for_linear_ce
yaox12 Feb 6, 2026
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
11 changes: 8 additions & 3 deletions megatron/core/model_parallel_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -246,9 +246,14 @@ class ModelParallelConfig:
Defaults to False.
"""

cross_entropy_fusion_impl: Literal['native', 'te'] = 'native'
"""If 'native', MCore based CE loss fusion is used, if 'te', Parallel CE loss
from Transformer Engine library is used. Defaults to 'native'.
cross_entropy_fusion_impl: Literal['native', 'te', 'linear'] = 'native'
"""
Specifies the implementation of cross-entropy loss fusion.

Options:
- 'native': Uses MCore-based cross-entropy loss fusion (default).
- 'te': Uses the parallel cross-entropy loss implementation from the Transformer Engine library.
- 'linear': Uses a linear-cross-entropy fusion approach.
"""

tp_comm_overlap_disable_qkv: bool = False
Expand Down
Loading