Low-Rank Curvature for Zeroth-Order Optimization in LLM Fine-tuning

초록

We introduce LOREN, a curvature-aware zeroth-order (ZO) optimization method for fine-tuning large language models (LLMs). Existing ZO methods, which estimate gradients via finite differences using random perturbations, often suffer from high variance and suboptimal search directions. Our approach addresses these challenges by: (i) reformulating the problem of gradient preconditioning as that of adaptively estimating an anisotropic perturbation distribution for gradient estimation, (ii) capturing curvature through a low-rank block diagonal preconditioner using the framework of natural evolution strategies, and (iii) applying a REINFORCEleave-one-out (RLOO) gradient estimator to reduce variance. Experiments on standard LLM benchmarks show that our method outperforms state-of-the-art ZOmethods by achieving higher accuracy and faster convergence, while cutting peak memory usage by up to 27.3% compared with MeZO-Adam.