This repository provides code examples of our CKA criterion for layer pruning, including some of our pruned models.
To observe and understand the functionality of our method, we simplify many training/fine-tuning parameters. If you are interested in reproducing our results, please follow the steps below:
1 - Put debug=false
2 - Increase the number of epochs in the fine-tuning function to 200.
3 - Divide the learning rate by 10 at epochs 100 and 150.
4 - Use data augmentation (details in the paper)
Please cite our paper in your publications if it helps your research.
@inproceedings{Pons:2024,
author = {Ian Pons,
Bruno Yamamoto,
Anna H. Reali Costa and
Artur Jordao},
title = {Effective Layer Pruning Through Similarity Metric Perspective},
booktitle = {International Conference on Pattern Recognition (ICPR).},
year = {2024},
}