Skip to content
Navigation Menu
Sign in
Products
Platform
GitHub
Repository hosting, code review, and collaboration
Copilot
AI coding agent for tasks, projects, and automated work
Build & Ship
Actions
CI/CD and automation
Code Quality
Quality analysis, fixes, and enforcement
Codespaces
Cloud development environments
Secure
Code Security
Vulnerability detection and remediation
Secret Protection
Secret scanning and leak prevention
All Features
Developers
Learn
Documentation
Training
Certifications
Build
Build with Copilot
Customize Copilot
Integrations
Explore
Trending repos
GitHub Sponsors
Developer articles
Changelog
Changelog
Your source for the latest features, improvements, and fixes across GitHub products, all in one place.
Businesses
By Stage
Enterprises
Team
Startups
Nonprofits
By Industry
Healthcare
Financial services
Manufacturing
Government
See all industries
Evaluate GitHub
Why GitHub
Customer Stories
Business insights
Trust Center
Add-ons
GitHub Advanced Security
Enterprise-grade code security
Copilot for business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Resources
Learn & Connect
Documentation
Resource center
Blog
Events & webinars
Community Forum
Customer Stories
Integrations
Support
Customer support
Training
GitHub Status
Company
About
Partners
Changelog
GitHub Universe
GitHub Universe
The event for developers, by developers — talks, workshops, and what’s next for GitHub.
Pricing
Search
/
Sign in
Sign up
Appearance settings
You signed in with another tab or window.
Reload
to refresh your session.
You signed out in another tab or window.
Reload
to refresh your session.
You switched accounts on another tab or window.
Reload
to refresh your session.
Dismiss alert
{{ message }}
NVIDIA
Megatron-LM
Repository navigation
Code
Issues
518
(518)
Pull requests
1.6k
(1.6k)
Discussions
Actions
Projects
Security and quality
Insights
More
items
All pull requests
Labels
104
(104)
Milestones
2
(2)
New pull request
Search pull requests
is
:
pr
state
:
open
is:pr state:open
Clear filter
Search
Pull requests
Open
1,622
(1,622)
Closed
5,753
(5,753)
Author
Label
Projects
Milestones
Reviews
Assignee
Sort by
Newest
descending
Sort…
Sort by Newest descending
More items
Comfortable display density
Compact display density
Enable per-head split-QKV Muon on the LayerShardedMuon NS home
#9571
·
wanyingw
opened
Nov 19, 2026
Contributor
·
7 tasks
·
·
1
Add element-granular dense buffer layout with an owner exchange
#9570
·
wanyingw
opened
Nov 19, 2026
Contributor
·
7 tasks
·
·
1
Gather the inference MoE unpermute through a slot map
complexity: medium
#9567
·
anlthms
opened
Nov 18, 2026
Contributor
·
1 task done
·
·
3
Batch five queued PRs: SwiGLU, Mamba, runtime CP, and inference serving
Batch PR
determinism-exempt
Run Merge-Queue jobs
#9566
·
balasaajay
opened
Nov 18, 2026
Contributor
·
·
2
build: Update TE to release_v2.22 and cuDNN frontend to 1.33
Disable selective unit tests
Run functional tests
#9565
·
balasaajay
opened
Nov 18, 2026
Contributor
·
3 tasks
·
·
5
Align parameter starts in the gradient buffer without the distributed optimizer
community-request
#9564
·
siddhant-rajhans
opened
Nov 18, 2026
·
5 of 7 tasks
·
·
3
Reduce-scatter sequence-parallel embeddings in the dynamic VLM forward
#9563
·
santhnm2
opened
Nov 18, 2026
Contributor
·
1 task
·
·
1
Keep the chat template's BOS when a request does not set add_BOS
community-request
#9562
·
siddhant-rajhans
opened
Nov 18, 2026
·
5 of 7 tasks
·
·
1
1
ci: Associate Codecov uploads with their pull requests
complexity: low
#9560
·
Connor-XY
opened
Nov 18, 2026
Contributor
·
1 task done
·
·
1
Fix QKV storage layout for static cached attention
#9559
·
mkhona-nvidia
opened
Nov 18, 2026
Contributor
·
2 of 4 tasks
·
·
1
Match vLLM's chat endpoint for reasoning fields and tool-call content
complexity: medium
#9558
·
santhnm2
opened
Nov 18, 2026
Contributor
·
4 of 7 tasks
·
·
2
Cast fp32 inputs in the inference-optimized fused norm + linear
complexity: low
Final Review
#9557
·
santhnm2
opened
Nov 18, 2026
Contributor
·
·
3
fix(ci): preserve the tested source tree with Testmon caching
complexity: low
#9556
·
wdykas
opened
Nov 18, 2026
Contributor
·
1 task
·
·
3
perf(models): avoid full-vocabulary copies in returned logits
Run NeMoRL tests
#9554
·
wdykas
opened
Nov 18, 2026
Contributor
·
1 task
·
·
2
Support eager MoK training with Megatron-FSDP v2
#9553
·
shjwudp
opened
Nov 18, 2026
Contributor
·
·
1
fix(GTP): keep gather-quantizer scales compact for NVFP4
complexity: low
#9552
·
xrennvidia
opened
Nov 18, 2026
Contributor
·
1 of 7 tasks
·
·
4
Save RNG state per data-parallel rank
#9551
·
dimapihtar
opened
Nov 18, 2026
Contributor
·
7 tasks
·
·
1
[2/2] Add an experimental vLLM streaming source to EAGLE3 finetuning
community-request
#9550
·
WEIXILIN0307
opened
Nov 18, 2026
·
·
1
[MFSDP v2] Prefetch for recomputation with a recorded execution order
complexity: low
#9549
·
XiaodaNV
opened
Nov 18, 2026
Contributor
·
1 task done
·
·
2
fix(moe): exclude padding routes from MoonEP expert placement
community-request
#9548
·
yyinter
opened
Nov 18, 2026
·
4 of 5 tasks
·
·
1
[MFSDP v2] Add bounded reduce-scatter windows and gradient view reuse
#9547
·
shjwudp
opened
Nov 18, 2026
Contributor
·
·
1
fix(moe): preserve BF16 expert inputs during activation offload
#9545
·
humairafirdowse18
opened
Nov 18, 2026
·
4 of 7 tasks
·
·
1
ci: Skip PR builds and tests for configured paths
#9544
·
balasaajay
opened
Nov 18, 2026
Contributor
·
1 task
·
·
1
docs: fix dead Energon and ModelOpt links in examples
community-request
#9542
·
farisbasha
opened
Nov 18, 2026
·
7 tasks
·
·
2
Check FSDP readiness before full CUDA graph capture
#9541
·
shjwudp
opened
Nov 18, 2026
Contributor
·
·
1
Previous
1
2
3
4
5
6
7
…
40
Next
You can’t perform that action at this time.