The rapid advancement of large-language models (LLMs) has driven extensive research into parameter compression after training has been completed, yet compression during the training phase remains largely unexplored. In this work, we introduce Rate-Constrained Training (BackSlash), a novel training-time compression approach based on rate-distortion optimization (RDO). BackSlash enables a flexible trade-off between model accuracy and complexity, significantly reducing parameter redundancy while preserving performance. Experiments in various architectures and tasks demonstrate that BackSlash can reduce memory usage by 60% - 90% without accuracy loss and provides significant compression gain compared to compression after training. Moreover, BackSlash proves to be highly versatile: it enhances generalization with small Lagrange multipliers, improves model robustness to pruning (maintaining accuracy even at 80% pruning rates), and enables network simplification for accelerated inference on edge devices.
View on arXiv@article{wu2025_2504.16968,
title={ BackSlash: Rate Constrained Optimized Training of Large Language Models },
author={ Jun Wu and Jiangtao Wen and Yuxing Han },
journal={arXiv preprint arXiv:2504.16968},
year={ 2025 }
}