Automatic Joint Structured Pruning and Quantization for Efficient Neural Network Training and Compression
Xiaoyi Qu, David Aponte, Colby Banbury, Daniel P. Robinson, Tianyu Ding, Kazuhito Koishida, Ilya Zharkov, Tianyi Chen, Microsoft, Lehigh University [email protected] , [email protected] · 2025
Read Automatic Joint Structured Pruning and Quantization for Efficient Neural Network Training and Compression by Xiaoyi Qu, David Aponte, Colby Banbury, Daniel P. Robinson, Tianyu Ding, Kazuhito Koishida, Ilya Zharkov, Tianyi Chen, Microsoft, Lehigh University [email protected] , [email protected] (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-16638.txt.
Description
Xiaoyi Qu and co-authors present a framework that shrinks neural networks by pruning and quantization at the same time, in a 2025 paper. Doing both together promises smaller models, but past schemes were fiddly, tuning-heavy, and narrow. Their framework, GETA, builds a dependency graph that maps which parts can be cut under quantization rules. A modified gradient method then guarantees each layer's bit budget is honored during training.
A joint learning strategy ties the two compressions together with interpretable links. On both image networks and transformers, GETA matched or beat existing joint methods. The process runs automatically on arbitrary architectures, which the authors stress is the point: compression should not need a bespoke pipeline per model. They target efficient deployment of large models on limited hardware.
Catalog record
| CREATOR | Xiaoyi Qu, David Aponte, Colby Banbury, Daniel P. Robinson, Tianyu Ding, Kazuhito Koishida, Ilya Zharkov, Tianyi Chen, Microsoft, Lehigh University [email protected] , [email protected] |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-16638