DxChat

AI Symptom Helper Chatbot 🧑‍⚕️

A Streamlit web application demonstrating an AI chatbot designed to provide general information about health symptoms. This tool uses open-source Large Language Models (LLMs) accessed via the Hugging Face Inference API.

➡️ Live Demo : [Link to the Streamlit Cloud App]

---

⚠️ EXTREMELY IMPORTANT DISCLAIMER ⚠️

• This is NOT a medical device or diagnostic tool.
• This application is for informational and demonstration purposes ONLY.
• The AI can make mistakes, provide inaccurate, incomplete, or biased information. It does NOT replace professional medical advice.
• NEVER use this tool to self-diagnose or make decisions about your health or treatment.
• ALWAYS consult a qualified healthcare provider for any health concerns, diagnosis, or treatment.
• If you think you have a medical emergency, call your doctor or emergency services IMMEDIATELY.
• Use of this tool is at your own risk. The developers assume no liability for any actions taken based on the information provided by this application.

---

Features

• Conversational Interface: Chat with an AI about health symptoms.
• Symptom Information: Receive general information potentially related to described symptoms (NOT a diagnosis).
• Open-Source LLM Powered: Utilizes models hosted on the Hugging Face Hub via their Inference API.
• Configurable: Easily change the underlying LLM model and system prompts via config.yaml.
• Safety Focused: Strong system prompts and disclaimers designed to prevent diagnosis and encourage professional consultation.
• Streamlit Cloud Ready: Designed for easy deployment on Streamlit Community Cloud.

Technology Stack

• Python: Core programming language.
• Streamlit: Web application framework.
• Hugging Face Inference API: Service to run LLMs remotely.
• huggingface_hub: Python client library for interacting with Hugging Face services.
• PyYAML: For handling the config.yaml configuration file.

Setup and Installation (Local Development)

1. Clone the Repository:

   git clone [(https://github.com/suhasaitham22/DxChat/tree/main)]
   cd [repository-folder-name]

2. Create a Virtual Environment:

   python -m venv venv
   # Activate the environment
   # On macOS/Linux:
   source venv/bin/activate
   # On Windows:
   .\venv\Scripts\activate

3. Install Dependencies:

   pip install -r requirements.txt

4. Hugging Face API Token:

   • You need a Hugging Face account: huggingface.co
   • Generate an API token with read permissions: huggingface.co/settings/tokens
   • Create a file named secrets.toml inside a .streamlit directory: .streamlit/secrets.toml
   • Add your token to the secrets.toml file:

     # .streamlit/secrets.toml
     [huggingface]
     token = "hf_YOUR_HUGGINGFACE_READ_TOKEN"

   • (Important): Add .streamlit/secrets.toml to your .gitignore file to avoid committing your secret token!

Running Locally

Once the setup is complete, run the Streamlit application:

streamlit run app.py