Skip to content

[bugfix] fix MTP checkpoint saving with TP > 1 #210

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
tastelikefeet merged 1 commit into from
Sep 25, 2026
Merged

[bugfix] fix MTP checkpoint saving with TP > 1 #210

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: 4 additions & 0 deletions src/mcore_bridge/model/modules/mtp_layer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,10 @@ def __init__(self, config: ModelConfig, submodules, *args, **kwargs):
if replace_eh_proj:
submodules.eh_proj = eh_proj
self.tp_group = getattr(self, 'tp_group', None)
if self.tp_group is None:
# mcore 0.16 doesn't set tp_group on the MTP layer. If it stays None, sharded_state_dict uses
# tp rank 0 on every TP rank and the replicated enorm/hnorm end up with two main replicas.
self.tp_group = parallel_state.get_tensor_model_parallel_group(check_initialized=False)
if not replace_eh_proj:
return
fp8_context = transformer_engine.pytorch.fp8_model_init(enabled=False)
Expand Down
Loading