Based on https://github.com/66RING/tiny-flash-attention.git
Blog：https://zhuanlan.zhihu.com/p/708867810
git submodule update --init --recursive
pip install .
python test.py
Benchmark on H20:
forked from weishengying/tiny-flash-attention
-
Notifications
You must be signed in to change notification settings - Fork 0
wangshang23/flash-attention2-cutlass
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
Flash-attention2 CUTLASS
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published
Languages
- Cuda 54.5%
- C++ 33.2%
- Python 12.3%