Abstract | ||
---|---|---|
Model compression has drawn great attention in deep learning community. A core problem in model compression is to determine the layer-wise optimal compression policy, e.g., the layer-wise bit-width in network quantization. Conventional hand-crafted heuristics rely on human experts and are usually sub-optimal, while recent reinforcement learning based approaches can be inefficient during the exploration of the policy space. In this article, we propose Bayesian automatic model compression (BAMC), which leverages non-parametric Bayesian methods to learn the optimal quantization bit-width for each layer of the network. BAMC is trained in a one-shot manner, avoiding the back and forth (re)-training in reinforcement learning based approaches. Experimental results on various datasets validate that our proposed methods can find reasonable quantization policies efficiently with little accuracy drop for the quantized network. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1109/JSTSP.2020.2977090 | IEEE Journal of Selected Topics in Signal Processing |
Keywords | DocType | Volume |
Bayesian learning,model compression,automatic machine learning,quantizartion,explainability | Journal | 14 |
Issue | ISSN | Citations |
4 | 1932-4553 | 2 |
PageRank | References | Authors |
0.36 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jiaxing Wang | 1 | 3 | 2.77 |
Haoli Bai | 2 | 18 | 5.62 |
Jiaxiang Wu | 3 | 2 | 0.36 |
Jian Cheng | 4 | 1327 | 115.72 |