-
Notifications
You must be signed in to change notification settings - Fork 3.9k
Pull requests: NVIDIA/Megatron-LM
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[PP] Add initial overlap_p2p_comm support for non-interleaved steady-state 1F1B
community-request
#4456
opened Apr 24, 2026 by
cky-dev
Loading…
Fix checkpoint loading with rerun state machine
Approved
All necessary approvals have been made
complexity: low
Add PackedSeqParams construction helpers.
community-request
#4447
opened Apr 23, 2026 by
bbuschkaemper
•
Draft
5 tasks done
Add functional tests for NVFP4 native param gather
community-request
#4446
opened Apr 23, 2026 by
re-imagined
Loading…
2 of 5 tasks
Handle SSM sharded tensor merge OOM with CPU fallback
community-request
#4442
opened Apr 23, 2026 by
returnL
Contributor
Loading…
2 of 5 tasks
Inference: Add the embedding and output layer in the full_iteration_inference cuda graph scope for hybrid models
complexity: low
Final Review
PR is in the "final review" stage
docs: fix broken links and anchors across READMEs and docs
docs-only
documentation only (docs or docstrings)
#4438
opened Apr 23, 2026 by
sbhavani
Contributor
Loading…
1 of 5 tasks
docs: fix Python version requirement and uv install commands in install
docs-only
documentation only (docs or docstrings)
#4437
opened Apr 23, 2026 by
sbhavani
Contributor
Loading…
5 tasks
get rid of weights_only=False
complexity: low
Expert Review
[deprecated] Apply this label to indicate that your PR is ready for expert review.
Final Review
PR is in the "final review" stage
Run functional tests
[Main] Fix invisible issues related to use_decoupled_grad for Megatron-FSDP.
complexity: low
Final Review
PR is in the "final review" stage
Previous Next
ProTip!
Exclude everything labeled
bug with -label:bug.