提出无需调参的零阶优化方法，高效微调大模型
Zero-order Parameter-free Optimization for LMO-based Methods: Novel Approach for Efficient Fine-tuning
- 结合零阶优化与非欧几何更新，自动适应参数
- 在13亿参数模型上实现稳定收敛，节省显存
- 适合资源受限场景下的大模型微调
大规模微调大语言模型（LLM）已成为现代优化的核心应用，使预训练模型能够适配多样下游任务和领域数据。大规模微调的主要障碍是反向传播带来的内存开销，需存储激活值、梯度和优化器状态。零阶（ZO）优化提供了一种内存高效的替代方案，但其性能对步长和平滑参数高度敏感，常需耗时的任务特定调优。无参数（PF）优化通过自适应算法参数解决此问题，无需依赖问题相关的常数。此外，大规模微调可受益于考虑参数块异质结构的几何感知更新，可通过利用线性最小化预言机（LMO）的方法建模。本文研究了基于LMO的零阶优化中的无参数自适应，并提出$ exttt{AdaNAGED}$方法，统一了无梯度训练、自适应调参与非欧更新几何。我们建立了收敛性保证，并在$ exttt{OPT}-1.3 ext{B}$模型的大规模微调任务上验证了该方法的有效性。
原文摘要 · Abstract (English)
Fine-tuning large language models (LLMs) has become a central application of modern optimization, enabling pretrained models to adapt to diverse downstream tasks and domain-specific data. A major obstacle in large-scale fine-tuning is the memory overhead of backpropagation, which requires storing activations, gradients, and optimizer states. Zeroth-order (ZO) optimization offers a memory-efficient alternative, but its performance is highly sensitive to the stepsize and smoothing parameter, often requiring costly task-specific tuning. Parameter-free (PF) optimization addresses this issue by adapting algorithmic parameters without prior knowledge of problem-dependent constants. Moreover, large-scale fine-tuning can benefit from geometry-aware updates that account for the heterogeneous structure of parameter blocks, which can be modeled through methods that exploit linear minimization oracle (LMO). In this work, we study PF adaptation for LMO-based ZO optimization and introduce $\texttt{AdaNAGED}$, a method that unifies gradient-free training, adaptive tuning, and non-Euclidean update geometry. We establish convergence guarantees and validate the method on large-scale LLM fine-tuning task with $\texttt{OPT}-1.3\mathrm{B}$ model.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。