Skip to content

Pull requests: NVIDIA/Megatron-LM

Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort

Pull requests list

[Dev] Add configurable attention latent normalization epsilon
#6204 opened Aug 3, 2026 by buptzyb Contributor Loading…
Fix GTP embedding prefetch buffer reuse ordering complexity: low nemotron
#6202 opened Aug 3, 2026 by JF-D Contributor Loading…
1 task done
Add inference state handoff support
#6196 opened Aug 1, 2026 by nvcsathe Contributor Draft
6 tasks
Add configurable server evaluation defaults
#6195 opened Aug 1, 2026 by nvcsathe Contributor Draft
6 tasks
Allocate main_grad on the reduce-scatter stream complexity: low Final Review PR is in the "final review" stage Run tests
#6187 opened Aug 1, 2026 by wujingyue Contributor Loading…
Document MFSDP optimizer design docs-only documentation only (docs or docstrings) Final Review PR is in the "final review" stage
#6186 opened Aug 1, 2026 by wujingyue Contributor Loading…
[Dev] Fix FSDP forward prefetch in combined 1F1B
#6174 opened Jul 31, 2026 by lhb8125 Contributor Draft
[Dev] Account for Qwen3.5 vision encoder FLOPs
#6173 opened Jul 31, 2026 by BestJuly Contributor Loading…
fix(ci): align workflow Python version with requires-python community-request waiting-on-maintainers Waiting on maintainers to respond
#6171 opened Jul 31, 2026 by MGPOCKY Loading…
3 tasks
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.