README.md

metadata

title: CGT-LLM-Beta RAG Chatbot
emoji: 🧬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 5.42.0
app_file: app.py
pinned: false
hf_oauth: true
hf_oauth_scopes:
  - inference-api
license: mit

🧬 CGT-LLM-Beta: Genetic Counseling RAG Chatbot

A Retrieval-Augmented Generation (RAG) chatbot for genetic counseling, cascade genetic testing, hereditary cancer syndromes, and related topics.

Features

• RAG System: Provides evidence-based answers from medical literature using vector database retrieval
• Multiple Models: Choose from various LLM models (Llama, Mistral, MediPhi, etc.)
• Education Level Adaptation: Answers are tailored to different education levels (Middle School, High School, College, Doctoral)
• Source Citations: View retrieved document chunks with similarity scores
• Readability Scoring: Flesch-Kincaid grade level scores for each answer
• OAuth Authentication: Secure access using Hugging Face OAuth tokens

How to Use

1. Log in: Click the "Login" button in the sidebar to authenticate with your Hugging Face account
2. Ask a question: Enter your question about genetic counseling, hereditary cancer, or related topics
3. Select options:
   • Choose your preferred LLM model
   • Select your education level for personalized answers
   • Adjust advanced settings (retrieval count, temperature, max tokens)
4. View results: See the answer, readability score, source documents, and similarity scores

Example Questions

The chatbot includes 100+ example questions covering topics like:

• BRCA1/BRCA2 mutations and cancer risk
• Lynch Syndrome (MLH1, MSH2, MSH6, PMS2, EPCAM)
• Genetic testing recommendations
• Family communication about genetic results
• Insurance and legal considerations (GINA)
• Screening and prevention strategies

Technical Details

• Vector Database: ChromaDB for fast semantic search
• Embeddings: Sentence-transformers (all-MiniLM-L6-v2)
• Inference: Hugging Face Inference API (via OAuth)
• Interface: Gradio 5.42.0

Important Notes

⚠️ Medical Disclaimer: This chatbot provides informational answers based on medical literature. It is not a substitute for professional medical advice, diagnosis, or treatment. Always consult with qualified healthcare providers for medical decisions.

Resources

The chatbot's knowledge base includes:

• NCCN Guidelines
• Medical literature on hereditary cancer syndromes
• Genetic counseling resources
• Patient education materials

---

Built with Gradio, huggingface_hub, and the Hugging Face Inference API.