混合优化：多任务微调对LLM性能的鸡尾酒效应——金融领域案例研究
Mixing It Up: The Cocktail Effect of Multi-Task Fine-Tuning on LLM Performance -- A Case Study in Finance
AI总结：
本研究以金融领域为案例，分析LLM微调策略，发现多任务微调（在相关任务组合上训练）比仅目标任务微调更有效，能让Phi-3-Mini等小模型超越GPT-4-o，还验证了通用指令数据正则化、数学数据迁移的作用。
摘要：
大型语言模型（LLMs）在金融等特定领域场景中的应用迅速扩展。特定领域LLMs通常基于其在该领域相关各类下游任务中的性能进行评估。本研究对LLMs针对此类任务的微调展开详细分析。与直觉略有相悖的是，我们发现在特定领域案例中，仅针对目标任务进行微调并非总是最有效的策略。相反，多任务微调——即模型在一系列相关任务的“鸡尾酒”组合上训练——能显著提升性能。我们展示了这种方法如何使Phi-3-Mini等小型模型取得最先进结果，甚至在金融基准测试中超越规模大得多的GPT-4-o模型。我们的研究涉及大规模实验，使用多个广泛采用的LLMs作为基线开展200多次训练实验，实证验证了多任务微调的优势。此外，我们探索将通用指令数据用作正则化形式，发现其有助于最小化性能退化。我们还研究了纳入数学数据的效果，发现数值推理能力的提升可有效迁移至金融任务。最后，我们指出，尽管针对下游任务的微调能带来任务性能的针对性提升，但不一定会带来领域知识或复杂领域推理能力的更广泛增益。
英文摘要：
The application of large language models (LLMs) in domain-specific contexts, including finance, has expanded rapidly. Domain-specific LLMs are typically evaluated based on their performance in various downstream tasks relevant to the domain. In this work, we present a detailed analysis of fine-tuning LLMs for such tasks. Somewhat counterintuitively, we find that in domain-specific cases, fine-tuning exclusively on the target task is not always the most effective strategy. Instead, multi-task finetuning - where models are trained on a cocktail of related tasks - can significantly enhance performance. We demonstrate how this approach enables a small model, such as Phi-3-Mini, to achieve state-of-the-art results, even surpassing the much larger GPT-4-o model on financial benchmarks. Our study involves a large-scale experiment, conducting over 200 training experiments using several widely adopted LLMs as baselines, and empirically confirms the benefits of multi-task fine-tuning. Additionally, we explore the use of general instruction data as a form of regularization, suggesting that it helps minimize performance degradation. We also investigate the inclusion of mathematical data, finding improvements in numerical reasoning that transfer effectively to financial tasks. Finally, we note that while fine-tuning for downstream tasks leads to targeted improvements in task performance, it does not necessarily result in broader gains in domain knowledge or complex domain reasoning abilities.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
