Skip to content

Navigation Menu

Explore
By company size
By use case
By industry
View all solutions
Topics
- AI
- DevOps
- Security
- Software Development
- View all
Explore
- GitHub Sponsors
  Fund open source developers
- The ReadME Project
  GitHub community articles
Repositories
- Enterprise platform
  AI-powered developer platform
Available add-ons
Pricing

Search code, repositories, users, issues, pull requests...

Search

Clear

Search syntax tips

Provide feedback

We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Saved searches

Use saved searches to filter your results more quickly

Name

Query

To see all available qualifiers, see our documentation.

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

karpathy / llm.c Public

Notifications You must be signed in to change notification settings
Fork 2.9k
Star 25.2k

Code
Issues 78
Pull requests 113
Discussions
Actions
Projects
Security
Insights

Additional navigation options

Code
Issues
Pull requests
Discussions
Actions
Projects
Security
Insights

cuDNN Forward Attention + FP16 non-cuDNN version in /dev/cuda/#215

Closed

ademeure wants to merge 8 commits intokarpathy:masterkarpathy/llm.c:masterfrom ademeure:forwardcudnnademeure/llm.c:forwardcudnnCopy head branch name to clipboard

+399-19

Conversation Commits8 (8)Checks Files changed

Commits

Commits on Apr 21, 2024

Initial working FP16 cuDNN Forward Attention (+ FP16 port of kernel 4)
ademeure
committed
Merge remote-tracking branch 'karpathy/master' into forwardcudnn
ademeure
committed
WIP - cuBLAS Ex() not working as expected...
ademeure
committed
Merge branch 'forwardcudnn' of https://github.com/ademeure/llm.c into forwardcudnn
ademeure
committed
Fix FP16/BF16 kernel version issues
ademeure
committed

Commits on Apr 22, 2024

Use -DENABLE_CUDNN for cuDNN path instead, and enable TF32 for attention_forward.cu by default
ademeure
committed
extra comments + tiny fix
ademeure
committed
remove unintentional train_gpu2.cu change
ademeure
committed

Footer

© 2025 GitHub, Inc.

Footer navigation

Terms
Privacy
Security
Status
Docs
Contact

You can’t perform that action at this time.