medical-o1-reasoning-SFT

## News [2025/04/22] We split the data and kept only the medical SFT dataset (`medical_o1_sft.json`). The file `medical_o1_sft_mix.json` contains a mix of medical and general instruction data. [2025/02/22] We released the [distilled dataset from Deepseek-R1](https://huggingface.co/datasets/FreedomIntelligence/Medical-R1-Distill-Data) based on medical verifiable problems. You can use it to initialize your models with the reasoning chain from `Deepseek-R1`. [2024/12/25] We open-sourced the medical reasoning dataset for SFT, built on medical verifiable problems and an LLM verifier. ## Introduction This dataset is used to fine-tune HuatuoGPT-o1, a medical LLM designed for advanced medical reasoning. This dataset is constructed using GPT-4o, which searches for solutions to [verifiable medical problems](https://huggingface.co/datasets/FreedomIntelligence/medical-o1-verifiable-problem) and validates them through a medical verifier. For details, see our [paper](https://arxiv.org/pdf/2412.18925) and [GitHub repository](https://github.com/FreedomIntelligence/HuatuoGPT-o1). ## Citation If you find our data useful, please consider citing our work! ``` @misc{chen2024huatuogpto1medicalcomplexreasoning, title={HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs}, author={Junying Chen and Zhenyang Cai and Ke Ji and Xidong Wang and Wanlong Liu and Rongsheng Wang and Jianye Hou and Benyou Wang}, year={2024}, eprint={2412.18925}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2412.18925}, } ```## 新闻 [2025/04/22] 我们对数据集进行了拆分，仅保留医疗监督微调（Supervised Fine-Tuning，SFT）数据集`medical_o1_sft.json`。文件`medical_o1_sft_mix.json`包含医疗与通用指令数据的混合数据集。 [2025/02/22] 我们基于医疗可验证问题，发布了源自Deepseek-R1的蒸馏数据集，其链接为https://huggingface.co/datasets/FreedomIntelligence/Medical-R1-Distill-Data。您可利用该数据集结合`Deepseek-R1`的推理链来初始化您的模型。 [2024/12/25] 我们开源了面向监督微调（SFT）的医疗推理数据集，该数据集基于医疗可验证问题与大语言模型（Large Language Model，LLM）验证器构建。 ## 简介 本数据集用于微调华佗GPT-o1——一款专为高级医疗推理设计的医疗大语言模型（Large Language Model，LLM）。本数据集依托GPT-4o构建，该模型会针对[医疗可验证问题](https://huggingface.co/datasets/FreedomIntelligence/medical-o1-verifiable-problem)搜索解决方案，并通过医疗验证器对方案进行校验。 如需了解更多细节，请参阅我们的[论文](https://arxiv.org/pdf/2412.18925)与[GitHub仓库](https://github.com/FreedomIntelligence/HuatuoGPT-o1)。 ## 引用 若您认为本数据集对您的研究有所帮助，请考虑引用我们的工作！ @misc{chen2024huatuogpto1medicalcomplexreasoning, title={HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs}, author={Junying Chen and Zhenyang Cai and Ke Ji and Xidong Wang and Wanlong Liu and Rongsheng Wang and Jianye Hou and Benyou Wang}, year={2024}, eprint={2412.18925}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2412.18925}, }

搜集汇总

背景与挑战

背景概述

该数据集是一个医学推理SFT数据集，用于微调HuatuoGPT-o1模型，以支持高级医学推理任务。它基于可验证的医学问题构建，使用GPT-4o生成解决方案并通过医学验证器验证，确保数据的准确性和专业性。数据集包含纯医学SFT数据和混合指令数据，源自Deepseek-R1的蒸馏数据，旨在提升模型的推理链能力。