Skip to content

Latest commit

 

History

History
13 lines (9 loc) · 540 Bytes

README.md

File metadata and controls

13 lines (9 loc) · 540 Bytes

ApproxConv: Approximating Continuous Convolutions for Deep Network Compression

This is the official code repository for the BMVC 2022 paper: "Approximating Continuous Convolutions for Deep Network Compression"

Requirements

Whilst an environment file is provided for use with ana/miniconda the core dependencies are:

- Python 3.8
- PyTorch 1.8.0
- PyTorch Lightning 1.5.3
- OmegaConf

You will also need to initialise the quantization submodule to run the quantisation experiments.