Fin-R1：一款通过强化学习实现金融推理的大型语言模型

Fin-R1: A Large Language Model for Financial Reasoning through Reinforcement Learning

March 20, 2025

作者: Zhaowei Liu, Xin Guo, Fangqi Lou, Lingfeng Zeng, Jinyi Niu, Zixuan Wang, Jiajie Xu, Weige Cai, Ziwei Yang, Xueqian Zhao, Chao Li, Sheng Xu, Dezhi Chen, Yun Chen, Zuo Bai, Liwen Zhang

cs.AI

摘要

推理型大型语言模型正在各个领域迅速发展。然而，它们在处理复杂金融任务方面的能力仍需深入探索。本文中，我们介绍了Fin-R1，一个专为金融领域设计的推理型大型语言模型。Fin-R1采用两阶段架构构建，利用基于DeepSeek-R1提炼和处理的金融推理数据集。通过监督微调（SFT）和强化学习（RL）训练，它在多种金融推理任务中展现了接近DeepSeek-R1的性能，参数规模为70亿。在我们的评估中，Fin-R1在FinQA和ConvFinQA任务上达到了同类LLM中的最先进水平（SOTA），并在其他任务中也超越了更大的模型。Fin-R1展示了强大的推理和决策能力，为金融领域遇到的各种问题提供了解决方案。我们的代码可在https://github.com/SUFE-AIFLM-Lab/Fin-R1获取。

English

Reasoning large language models are rapidly evolving across various domains. However, their capabilities in handling complex financial tasks still require in-depth exploration. In this paper, we introduce Fin-R1, a reasoning large language model specifically designed for the financial sector. Fin-R1 is built using a two-stage architecture, leveraging a financial reasoning dataset distilled and processed based on DeepSeek-R1. Through supervised fine-tuning (SFT) and reinforcement learning (RL) training, it demonstrates performance close to DeepSeek-R1 with a parameter size of 7 billion across a range of financial reasoning tasks. It achieves the state-of-the-art (SOTA) in the FinQA and ConvFinQA tasks between those LLMs in our evaluation, surpassing larger models in other tasks as well. Fin-R1 showcases strong reasoning and decision-making capabilities, providing solutions to various problems encountered in the financial domain. Our code is available at https://github.com/SUFE-AIFLM-Lab/Fin-R1.