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
Accelerator
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
Type
/
to 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
800
Star
3.5k
Code
Issues
159
Pull requests
164
Discussions
Actions
Security and quality
0
Insights
Additional navigation options
Code
Issues
Pull requests
Discussions
Actions
Security and quality
Insights
[PyTorch] Preserve FP8 recompute state for inner autocast
- #3284
#3284
Merged
pggPL
merged 13 commits into
NVIDIA:main
NVIDIA/TransformerEngine:main
from
nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen/TransformerEngine:fix/fp8-inner-autocast-recompute
Copy head branch name to clipboard
Aug 10, 2026
Conversation
Commits
13
(13)
Checks
Files changed
Merged
[PyTorch] Preserve FP8 recompute state for inner autocast
#3284
pggPL
merged 13 commits into
NVIDIA:main
NVIDIA/TransformerEngine:main
from
nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen/TransformerEngine:fix/fp8-inner-autocast-recompute
Copy head branch name to clipboard
Commits
Commits on Aug 2, 2026
fix: preserve FP8 recompute state for inner autocast
Show description for 317e256
nvegesna-netizen
committed
317e256
View commit details
Copy full SHA for 317e256
Browse repository at this point
test: fix activation recompute test license header
Show description for 4fb02b0
nvegesna-netizen
committed
4fb02b0
View commit details
Copy full SHA for 4fb02b0
Browse repository at this point
test: run FP8 recompute coverage in PyTorch QA
Show description for 34cf656
nvegesna-netizen
committed
34cf656
View commit details
Copy full SHA for 34cf656
Browse repository at this point
test: compare inner FP8 autocast recompute numerics
Show description for 00c727f
nvegesna-netizen
committed
00c727f
View commit details
Copy full SHA for 00c727f
Browse repository at this point
Commits on Aug 4, 2026
refactor: rename FP8 recompute flag to _IN_ACTIVATION_RECOMPUTE_REGION
Show description for 281e97f
nvegesna-netizen
committed
281e97f
View commit details
Copy full SHA for 281e97f
Browse repository at this point
refactor: rename recompute phase global for symmetry
Show description for 199aa6f
nvegesna-netizen
committed
199aa6f
View commit details
Copy full SHA for 199aa6f
Browse repository at this point
test: restore dropped FP8 recompute coverage and assert stash bookkeeping
Show description for 3e4c296
nvegesna-netizen
committed
3e4c296
View commit details
Copy full SHA for 3e4c296
Browse repository at this point
test: make recompute stash assertions meaningful, guard bf16
Show description for 2ce745f
nvegesna-netizen
committed
2ce745f
View commit details
Copy full SHA for 2ce745f
Browse repository at this point
test: assert the recompute reinstalls the stashed FP8 state
Show description for 9902b53
nvegesna-netizen
committed
9902b53
View commit details
Copy full SHA for 9902b53
Browse repository at this point
test: drop tautological recompute state assertions
Show description for 0d2e9b2
nvegesna-netizen
committed
0d2e9b2
View commit details
Copy full SHA for 0d2e9b2
Browse repository at this point
test: pin the changed predicate directly
Show description for 4c031dc
nvegesna-netizen
committed
4c031dc
View commit details
Copy full SHA for 4c031dc
Browse repository at this point
Merge branch 'main' into fix/fp8-inner-autocast-recompute
nvegesna-netizen
authored
14ba6a2
View commit details
Copy full SHA for 14ba6a2
Browse repository at this point
Commits on Aug 5, 2026
Merge branch 'main' into fix/fp8-inner-autocast-recompute
nvegesna-netizen
authored
fd58f6b
View commit details
Copy full SHA for fd58f6b
Browse repository at this point
You can’t perform that action at this time.
