Enhancing Split Computing and Early Exit Applications through Predefined Sparsity

Official implementation of the paper Enhancing Split Computing and Early Exit Applications through Predefined Sparsity accepted at the 27th Forum on specification and Design Languages (FDL 2024).

Installation

1. Repository setup:

• $ git clone https://github.com/intelligolabs/sparsity_sc_ee.git
• $ cd sparsity_sc_ee
• $ mv mnist.npz mnist/

2. Conda environment setup:

• $ conda create -n sparsity_sc_ee python=3.8
• $ conda activate sparsity_sc_ee
• $ pip install -r requirements.txt

Run Sparsity SC&EE

To run Sparsity SC&EE, use the file main.py. In particular, the launch.sh file contains two launch script examples that you can use to modify the default configuration.

Credits

We would like to thank Sourya Dey for the repository predefinedsparse-nnets.

Authors

Luigi Capogrosso¹, Enrico Fraccaroli^1,2, Giulio Petrozziello³, Francesco Setti¹, Samarjit Chakraborty², Franco Fummi¹, Marco Cristani¹

¹ Department of Engineering for Innovation Medicine, University of Verona, Italy

² Department of Computer Science, The University of North Carolina at Chapel Hill, USA

³ Department of Computer Science, University of Verona, Italy

Citation

If you use Sparsity SC&EE, please, cite the following paper:

@article{capogrosso2024enhancing,
  title={Enhancing Split Computing and Early Exit Applications through Predefined Sparsity},
  author={Capogrosso, Luigi and Fraccaroli, Enrico and Petrozziello, Giulio and Setti, Francesco and Chakraborty, Samarjit and Fummi, Franco and Cristani, Marco},
  journal={arXiv preprint arXiv:2407.11763},
  year={2024}
}