Skip to content

Fix checkpoint converter missing parallel group initialization #3217

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
dimapihtar merged 1 commit into from
Feb 4, 2026
Merged

Fix checkpoint converter missing parallel group initialization #3217

Show file tree
Hide file tree
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
2 changes: 1 addition & 1 deletion pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,7 @@ Download = "https://github.com/NVIDIA/Megatron-LM/releases"
Homepage = "https://github.com/NVIDIA/Megatron-LM/megatron/core"

[project.optional-dependencies]
mlm = ["flask-restful", "sentencepiece", "tiktoken", "wandb", "transformers"]
mlm = ["flask-restful", "sentencepiece", "tiktoken", "wandb", "transformers", "accelerate"]

dev = [
"nvidia-modelopt[torch]; sys_platform != 'darwin'",
Expand Down
6 changes: 6 additions & 0 deletions tools/checkpoint/saver_base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -170,9 +170,15 @@ def initialize_megatron_env(self):

# For backward compatibility during local parallel states refactoring
fake_tp_group = _ConverterFakeProcessGroup(size=self.args.target_tensor_parallel_size)
fake_pp_group = _ConverterFakeProcessGroup(size=self.args.target_pipeline_parallel_size)
fake_ep_group = _ConverterFakeProcessGroup(size=self.args.target_expert_parallel_size)
fake_dp_group = _ConverterFakeProcessGroup(size=1)
mpu._TENSOR_MODEL_PARALLEL_GROUP = fake_tp_group
mpu._PIPELINE_MODEL_PARALLEL_GROUP = fake_pp_group
mpu._EXPERT_MODEL_PARALLEL_GROUP = fake_ep_group
mpu._DATA_PARALLEL_GROUP = fake_dp_group
mpu._DATA_PARALLEL_GROUP_WITH_CP = fake_dp_group
mpu._INTRA_PARTIAL_DATA_PARALLEL_GROUP_WITH_CP = fake_dp_group
fused_kernels.load(self.margs)

try:
Expand Down
65 changes: 43 additions & 22 deletions uv.lock
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

Loading