Expressive and Generalizable Low-rank Adaptation for Large Models via Slow Cascaded Learning
This addresses the problem of efficient and robust fine-tuning for large models, offering an incremental improvement over existing LoRA variants.
The paper tackles the limited expressiveness, overfitting, and hyperparameter sensitivity of low-rank adaptation (LoRA) for fine-tuning large models by introducing LoRA Slow Cascade Learning (LoRASC), which significantly outperforms existing baselines on various language and vision datasets while improving stability and out-of-distribution robustness.
Efficient fine-tuning plays a fundamental role in modern large models, with low-rank adaptation emerging as a particularly promising approach. However, the existing variants of LoRA are hampered by limited expressiveness, a tendency to overfit, and sensitivity to hyperparameter settings. This paper presents LoRA Slow Cascade Learning (LoRASC), an innovative technique designed to enhance LoRA's expressiveness and generalization capabilities while preserving its training efficiency. Our approach augments expressiveness through a cascaded learning strategy that enables a mixture-of-low-rank adaptation, thereby increasing the model's ability to capture complex patterns. Additionally, we introduce a slow-fast update mechanism and cascading noisy tuning to bolster generalization. The extensive experiments on various language and vision datasets, as well as robustness benchmarks, demonstrate that the proposed method not only significantly outperforms existing baselines, but also mitigates overfitting, enhances model stability, and improves OOD robustness. Code will be release in https://github.com/microsoft/LoRASC very soon.