Skip to content
Navigation Menu
Sign in
Appearance settings
Platform
AI CODE CREATION
GitHub Copilot
Write better code with AI
GitHub Copilot app
Direct agents from issue to merge
MCP Registry
Integrate external tools
DEVELOPER WORKFLOWS
Actions
Automate any workflow
Codespaces
Instant dev environments
Issues
Plan and track work
Code Review
Manage code changes
Code Quality
Enforce quality at merge
APPLICATION SECURITY
GitHub Advanced Security
Find and fix vulnerabilities
Code security
Secure your code as you build
Secret protection
Stop leaks before they start
EXPLORE
Why GitHub
Documentation
Blog
Changelog
Marketplace
View all features
Solutions
BY COMPANY SIZE
Enterprises
Small and medium teams
Startups
Nonprofits
BY USE CASE
App Modernization
DevSecOps
DevOps
CI/CD
View all use cases
BY INDUSTRY
Healthcare
Financial services
Manufacturing
Government
View all industries
View all solutions
Resources
EXPLORE BY TOPIC
AI
Software Development
DevOps
Security
View all topics
EXPLORE BY TYPE
Customer stories
Events & webinars
Ebooks & reports
Business insights
GitHub Skills
SUPPORT & SERVICES
Documentation
Customer support
Community forum
Trust center
Partners
View all resources
Open Source
COMMUNITY
GitHub Sponsors
Fund open source developers
PROGRAMS
Security Lab
Maintainer Community
GitHub Stars
Archive Program
REPOSITORIES
Topics
Trending
Collections
Enterprise
ENTERPRISE SOLUTIONS
Enterprise platform
AI-powered developer platform
AVAILABLE ADD-ONS
GitHub Advanced Security
Enterprise-grade security features
Copilot for Business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
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 }}
Uh oh!
There was an error while loading.
Please reload this page
.
NVIDIA
/
TransformerEngine
Public
Notifications
You must be signed in to change notification settings
Fork
811
Star
3.5k
Code
Issues
159
Pull requests
186
Discussions
Actions
Security and quality
0
Insights
Additional navigation options
Code
Issues
Pull requests
Discussions
Actions
Security and quality
Insights
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel
- #3315
#3315
Merged
Oleg-Goncharov
merged 20 commits into
NVIDIA:main
NVIDIA/TransformerEngine:main
from
cael-ling:feature/mxfp8-group-swiglu-recompute
cael-ling/TransformerEngine:feature/mxfp8-group-swiglu-recompute
Copy head branch name to clipboard
Aug 28, 2026
Conversation
Commits
20
(20)
Checks
Files changed
Merged
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel
#3315
Oleg-Goncharov
merged 20 commits into
NVIDIA:main
NVIDIA/TransformerEngine:main
from
cael-ling:feature/mxfp8-group-swiglu-recompute
cael-ling/TransformerEngine:feature/mxfp8-group-swiglu-recompute
Copy head branch name to clipboard
Commits
Commits on Aug 4, 2026
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel
Show description for 5ae2ae8
cael-ling
committed
5ae2ae8
View commit details
Copy full SHA for 5ae2ae8
Browse repository at this point
[pre-commit.ci] auto fixes from pre-commit.com hooks
Show description for 2979881
pre-commit-ci[bot]
committed
2979881
View commit details
Copy full SHA for 2979881
Browse repository at this point
Commits on Aug 5, 2026
[PyTorch] Require contiguous, colocated operands in group_swiglu_quantize
Show description for e092890
cael-ling
committed
e092890
View commit details
Copy full SHA for e092890
Browse repository at this point
[PyTorch] Require grouped metadata tensors on the current CUDA device
Show description for 5bfd791
cael-ling
committed
5bfd791
View commit details
Copy full SHA for 5bfd791
Browse repository at this point
[PyTorch] Scope the grouped metadata device check to group_swiglu_quantize
Show description for 8572e3a
cael-ling
committed
8572e3a
View commit details
Copy full SHA for 8572e3a
Browse repository at this point
Commits on Aug 8, 2026
[Common/PyTorch] Rename to group_scaled_swiglu and speed up the kernel
Show description for a5f6859
cael-ling
committed
a5f6859
View commit details
Copy full SHA for a5f6859
Browse repository at this point
[PyTorch] Add a benchmark for group_scaled_swiglu
Show description for 0b94fcb
cael-ling
committed
0b94fcb
View commit details
Copy full SHA for 0b94fcb
Browse repository at this point
Commits on Aug 19, 2026
[Common/PyTorch] Halve the silu MUFU cost and add clamped scaled SwiGLU
Show description for 947761e
cael-ling
committed
947761e
View commit details
Copy full SHA for 947761e
Browse repository at this point
[Common] Cut the clamped scaled SwiGLU ALU cost
Show description for 5f1452e
cael-ling
committed
5f1452e
View commit details
Copy full SHA for 5f1452e
Browse repository at this point
[PyTorch] Benchmark the clamped kernel against a clamped baseline
Show description for 8ff9917
cael-ling
committed
8ff9917
View commit details
Copy full SHA for 8ff9917
Browse repository at this point
Commits on Aug 21, 2026
Merge branch 'main' into feature/mxfp8-group-swiglu-recompute
Show description for f0d1610
vthumbe1503
authored
f0d1610
View commit details
Copy full SHA for f0d1610
Browse repository at this point
[pre-commit.ci] auto fixes from pre-commit.com hooks
Show description for dc56317
pre-commit-ci[bot]
committed
dc56317
View commit details
Copy full SHA for dc56317
Browse repository at this point
Update transformer_engine/pytorch/csrc/extensions/cast.cpp
Show description for 08866bc
vthumbe1503
and
greptile-apps[bot]
authored
08866bc
View commit details
Copy full SHA for 08866bc
Browse repository at this point
Commits on Aug 25, 2026
[Common][PyTorch] Promote the approximate SwiGLU operators to math.h
Show description for ed28bf0
cael-ling
committed
ed28bf0
View commit details
Copy full SHA for ed28bf0
Browse repository at this point
[pre-commit.ci] auto fixes from pre-commit.com hooks
Show description for e95cf87
pre-commit-ci[bot]
committed
e95cf87
View commit details
Copy full SHA for e95cf87
Browse repository at this point
Merge branch 'main' into feature/mxfp8-group-swiglu-recompute
vthumbe1503
authored
5eab1f3
View commit details
Copy full SHA for 5eab1f3
Browse repository at this point
Commits on Aug 26, 2026
[PyTorch] Check the grouped scaled SwiGLU against TE's SwiGLU reference
Show description for 66aa446
cael-ling
committed
66aa446
View commit details
Copy full SHA for 66aa446
Browse repository at this point
Merge branch 'main' into feature/mxfp8-group-swiglu-recompute
Oleg-Goncharov
authored
37fd277
View commit details
Copy full SHA for 37fd277
Browse repository at this point
Commits on Aug 27, 2026
Merge branch 'main' into feature/mxfp8-group-swiglu-recompute
Oleg-Goncharov
authored
a585043
View commit details
Copy full SHA for a585043
Browse repository at this point
Commits on Aug 28, 2026
Merge branch 'main' into feature/mxfp8-group-swiglu-recompute
Oleg-Goncharov
authored
58d0b2e
View commit details
Copy full SHA for 58d0b2e
Browse repository at this point
You can’t perform that action at this time.
![pre-commit-ci[bot]](https://avatars.githubusercontent.com/in/68672?v=4&size=32){kind=avatar}
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?v=4&size=32){kind=avatar}
