---
license: afl-3.0
language:
- en
- zh
metrics:
- accuracy
base_model:
- deepseek-ai/DeepSeek-R1-Distill-Llama-70B
pipeline_tag: text-generation
library_name: transformers
tags:
- medical
- deepseek-r1
- health
- ehr
- reasoning

# 核心配置：开启门控
# gated: true

# （可选）申请表单上方的提示语
# extra_gated_heading: "Access Request"
# extra_gated_description: "Please provide your organization and intended use."

# （核心）自定义问卷字段
#extra_gated_fields:
#  Affiliation: text          # 单位/机构（文本框，满足你的需求）
#  Research Purpose: text     # 研究用途（文本框）
#  Country: text              # 国家（文本框）
---

<div align="center">
  <h1>🧬 RareSeek-R1</h1>
  <h3>A Specialized Language Model for Rare Disease Diagnosis and Clinical Reasoning</h3>
</div>

<p align="center">
  <img src="https://img.shields.io/badge/Language-English%20%7C%20Chinese-blue"> <img src="https://img.shields.io/badge/Task-Clinical_Reasoning-brightgreen">
  <a href="https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B"> <img src="https://img.shields.io/badge/Base_Model-DeepSeek--R1--Distill--Llama--70B-orange"></a>
  <a href="https://opensource.org/licenses/AFL-3.0"><img src="https://img.shields.io/badge/License-AFL--3.0-gray"></a>
</p>

## 📖 Model Overview
**RareSeek-R1** is a domain-specialized large language model tailored for rare-disease diagnostic reasoning, developed through a Progressive Parameter-Efficient Transfer Learning framework. 

The model is first instruction-tuned on the clinically grounded **RareMed-Corpus**, a large, multi-source dataset deeply integrated from medical textbooks, guidelines, biomedical literature, and real-world EHR (Electronic Health Record) narratives. It is subsequently fine-tuned on **RareMed-CoT**, a high-fidelity corpus designed to instill explicit, stepwise clinical reasoning that aligns with real-world diagnostic workflows.

<p align="center">
  <img src="https://github.com/mulinlab/RareSeek-R1/raw/main/RareSeek-R1.png" alt="Figure 1: Overall framework and pipeline of RareSeek-R1." width="800">
  <em>Figure 1: Overall framework and pipeline of RareSeek-R1.</em>
</p>

## 🗄️ Dataset & Resources
📚 **RareMedData**: Access the comprehensive medical dataset used for training here:  
👉 [https://huggingface.co/datasets/TaoMedAI/RareMedData](https://huggingface.co/datasets/TaoMedAI/RareMedData)