Chemistry

<div align="center">

AI Agents in Chemical Research: GVIM - An Intelligent Research Assistant System 🧪🤖

English | 简体中文

An intelligent research assistant system designed specifically for chemical science, featuring fine-tuned language models and specialized chemistry capabilities.

</div>

🌟 Highlights

<table> <tr> <td>

🧬 Core Features

• Fine-tuned LLMs for chemistry
• Molecular visualization
• Literature retrieval & analysis
• Multimodal capabilities

</td> <td>

🚀 Key Benefits

• Specialized for chemical research
• Continuous learning ability
• Free GPU resources on Colab
• Comprehensive documentation

</td> </tr> </table>

📚 Model Overview

GVIM combines several cutting-edge technologies:

🔬 Technical Features

• Fine-tuned Models: Trained on curated chemistry instruction data
• Chemistry Tools: Molecular visualization and SMILES processing
• Smart Retrieval: Advanced chemical literature search and analysis
• Multimodal Support: Formula recognition and image analysis
• Knowledge Base: Local document processing and continuous learning

🎯 Main Applications

1. Chemical research assistance and analysis
2. Molecular structure visualization
3. Literature review and knowledge extraction
4. Handwritten formula recognition
5. Document-based knowledge processing

💻 Quick Start

Requirements

# Create conda environment
conda create -n gvim python=3.9.19
conda activate gvim

# Install dependencies
pip install -r requirements.txt

API Configuration ⚙️

Required API keys:

TAVILY_API_KEY="your_key_here"
REPLICATE_API_TOKEN="your_token_here"
Groq_API_KEY="your_key_here"

Launch 🚀

python app.py

🎥 Key Features Showcase

<details> <summary>📸 Click to view feature demonstrations</summary>

1. Nature Chemistry Search Interface

2. Multimodal Chemical Recognition

3. Local Document Analysis

</details>

🆓 Free GPU Resources

Access GVIM's capabilities using Colab's free GPU resources:

📖 Citation

@article{Digital Discovery,
  author = {Kangyong Ma},
  affiliation = {College of Physics and Electronic Information Engineering, Zhejiang Normal University},
  address = {Jinhua City, 321000, China},
  doi = {10.1039/D4DD00398E},
  email = {[email protected], [email protected]}
}

📞 Contact

<table> <tr> <td>

📧 Email

• [email protected]
• [email protected]

</td> <td>

🏢 Institution

College of Physics and Electronic Information Engineering
Zhejiang Normal University
Jinhua City, 321000, China

</td> </tr> </table>

---

<div align="center">

🌟 Star us on GitHub | 📌 Commercial use requires authorization | 📑 Read our paper

</div>