Instructions to use kernels-community/flash-attn2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Kernels
How to use kernels-community/flash-attn2 with Kernels:
# !pip install kernels from kernels import get_kernel kernel = get_kernel("kernels-community/flash-attn2") - Notebooks
- Google Colab
- Kaggle
Would it be possible to maintain the torch 2.6 builds for longer?
#1
by hbenoit - opened
axolotl currently defaults to torch 2.6.0, and getting flash-attn to run correctly is proving quite tricky on my setup (getting obscure symbol errors). So defaulting to the kernels builds has been quite useful. However, the code that ran yesterday doesn't run anymore because the builds were removed :)