Skip to content

Pull requests: NVIDIA/Megatron-LM

New pull request New
239 Open 975 Closed
239 Open 975 Closed
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
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

update checkpointing documentation
#2606 opened Dec 9, 2025 by dimapihtar Loading…
6 tasks
Core 0.16
2
cp: [Was PR1912][Dev] feat(moe): Fine-grained activation offloading (#1969) into core_dev_r0.15.0
#2605 opened Dec 9, 2025 by chtruong814 Loading…
6 tasks
1
Use FP4 context for mamba
#2604 opened Dec 9, 2025 by kwyss-nvidia Loading…
6 tasks
Core 0.16
2
Modify config for the PP fix. Expert Review Apply this label to indicate that your PR is ready for expert review.
#2603 opened Dec 9, 2025 by yobibyte Loading…
6 tasks
Core 0.16
ci: Pin gojq
#2602 opened Dec 9, 2025 by ko3n1g Loading…
6 tasks
Core 0.16
Inference | Add request only if no paused requests. Expert Review Apply this label to indicate that your PR is ready for expert review.
#2600 opened Dec 9, 2025 by lmcafee-nvidia Loading…
6 tasks
Core 0.15
@lmcafee-nvidia
NVLS - fused reduce-scatter + residual + rms-norm + all-gather kernel
#2599 opened Dec 9, 2025 by sidsingh-nvidia Loading…
5 of 6 tasks
Core 0.16
1
Dnarayanan/latent moe
#2594 opened Dec 9, 2025 by pablo-garay Loading…
6 tasks
Core 0.16
2
Check skip_prompt_log_probs in add_request Expert Review Apply this label to indicate that your PR is ready for expert review.
#2593 opened Dec 9, 2025 by tdene Loading…
6 tasks
Core 0.16
[docs] Create blank docs framework
#2592 opened Dec 9, 2025 by Phlip79 • Draft
6 tasks
1
added pad tokens as dummy sequence
#2591 opened Dec 8, 2025 by jalbericiola Loading…
6 tasks
Core 0.16
2
Explicitly zero out padding token outputs when using quantization scales
#2585 opened Dec 8, 2025 by santhnm2 • Draft
6 tasks
1
Inference | Fix entangled request generations. Expert Review Apply this label to indicate that your PR is ready for expert review.
#2584 opened Dec 8, 2025 by lmcafee-nvidia Loading…
6 tasks
Core 0.15
@lmcafee-nvidia
2
Ignore log level in functional test
#2579 opened Dec 5, 2025 by kwyss-nvidia Loading…
6 tasks
1
Synchronize total block count across pipeline parallel ranks
#2578 opened Dec 5, 2025 by santhnm2 Loading…
6 tasks
2
fix: ckpt loading failed because of padding metadata in dist optimizer Expert Review Apply this label to indicate that your PR is ready for expert review.
#2576 opened Dec 5, 2025 by yaoyu-33 Loading…
6 tasks
@yaoyu-33
2
[Megatron-FSDP] Support both old and new DeviceMesh APIs. Expert Review Apply this label to indicate that your PR is ready for expert review.
#2575 opened Dec 5, 2025 by cspades Loading…
3 of 6 tasks
Core 0.16
@shjwudp @cspades
1
Add initial support for Kimi Delta Attention (Feature Request #2446) community-request
#2573 opened Dec 5, 2025 by CodersAcademy006 Loading…
1
15
partial cudagraph scopes and improvements for training
#2572 opened Dec 5, 2025 by jiemingz Loading…
6 tasks
@jiemingz
1
Fix: Ensure token IDs respect vocab_size in dataset, embeddings, and … community-request
#2570 opened Dec 5, 2025 by CodersAcademy006 Loading…
1
6
[Dev] Improve MoE Logging
#2569 opened Dec 5, 2025 by yanring • Draft
6 tasks
Core 0.16
2
[Dev] Optimize TE CUDA Graph _get_sample_arguments() Time
#2568 opened Dec 5, 2025 by buptzyb Loading…
6 tasks
Core 0.16
@buptzyb
4
Add offset method for slow tokenizer community-request
#2567 opened Dec 5, 2025 by cael-ling Loading…
6 tasks
1
feat: Api compat add decorator dev
#2545 opened Dec 4, 2025 by pablo-garay Loading…
6 tasks
1
[docs] Use autodoc2 and remove automodule
#2542 opened Dec 4, 2025 by Phlip79 • Draft
6 tasks
1
ProTip! Exclude everything labeled bug with -label:bug.