Metadata-Version: 2.1
Name: power-attention-cuda
Version: 0.1.0
Summary: CUDA implementation of power attention
Requires-Python: >=3.11
Description-Content-Type: text/markdown
Requires-Dist: torch>=2.6.0
Requires-Dist: numpy>=2.2

This is the legacy CUDA kernel for Power Attention, refer to https://github.com/m-a-n-i-f-e-s-t/power-attention instead.


## Build

Build a single wheel

```bash
./build_wheels.sh -t <torch version> -p <python version>
```

Build all wheels 

```bash
./build_all_wheels.sh
```
