$\beta$-DPO: Direct Preference Optimization with Dynamic $\beta$

TL;DR

This paper introduces a dynamic $eta$ calibration method for Direct Preference Optimization, improving LLM alignment with human preferences by adapting to data quality and filtering out outliers.

Contribution

It proposes a novel framework that dynamically adjusts $eta$ during training and incorporates data filtering, enhancing DPO's robustness and performance.

Findings

Dynamic $eta$ improves model alignment with preferences.

Data filtering reduces the impact of outliers.

Significant performance gains across models and datasets.

Abstract

Direct Preference Optimization (DPO) has emerged as a compelling approach for training Large Language Models (LLMs) to adhere to human preferences. However, the performance of DPO is sensitive to the fine-tuning of its trade-off parameter $\beta$, as well as to the quality of the preference data. We analyze the impact of $\beta$ and data quality on DPO, uncovering that optimal $\beta$ values vary with the informativeness of pairwise data. Addressing the limitations of static $\beta$ values, we introduce a novel framework that dynamically calibrates $\beta$ at the batch level, informed by data quality considerations. Additionally, our method incorporates $\beta$-guided data filtering to safeguard against the influence of outliers. Through empirical evaluation, we demonstrate that our dynamic $\beta$ adjustment technique significantly improves DPO's performance across a range of models and datasets, offering a more robust and adaptable training paradigm for aligning LLMs with human feedback. The code is available at \url{https://github.com/junkangwu/beta-DPO}.