Metadata-Version: 2.4
Name: mm-kermac
Version: 0.1.1
Summary: Pytorch routines for (Ker)nel (Mac)hines
Project-URL: Homepage, https://github.com/MetaMachines/mm-kermac-py
Project-URL: Repository, https://github.com/MetaMachines/mm-kermac-py.git
Author-email: MetaMachines LLC <contact@metamachines.co>
License: MIT
License-File: LICENSE
Requires-Python: >=3.8
Requires-Dist: cuda-toolkit[cccl]
Requires-Dist: mm-ptx>=0.1.2
Requires-Dist: numpy>=1.21
Requires-Dist: platformdirs>=3.0
Requires-Dist: torch>=2.0
Provides-Extra: cu11
Requires-Dist: cuda-core[cu11]>=0.3.2; extra == 'cu11'
Provides-Extra: cu12
Requires-Dist: cuda-core[cu12]>=0.3.2; extra == 'cu12'
Provides-Extra: cu13
Requires-Dist: cuda-core[cu13]>=0.3.2; extra == 'cu13'
Description-Content-Type: text/markdown

# mm-kermac
> Dynamically compiled hyper semirings for Pytorch using PTX Inject and Stack PTX

## Installation
```bash
pip install mm-kermac
```

## TODO
    * Better Docs
    * Benchmarks

## License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.

## Citation
If you use this software in your work, please cite it using the following BibTeX entry (generated from the [CITATION.cff](CITATION.cff) file):
```bibtex
@software{Durham_mm-kermac_2025,
  author       = {Durham, Charlie},
  title        = {mm-kermac: PTX Inject and Stack PTX for Python},
  version      = {0.1.1},
  date-released = {2025-10-19},
  url          = {https://github.com/MetaMachines/mm-kermac-py}
}
```