QnA ChatBot using Langchain 🔗 and Streamlit

A mini project that showcases a QnA-style LLM application built with Streamlit for UI and the Langchain ecosystem for LLM orchestration.

Index:

• Project Details
  • Aim
  • Features
  • Tech Stack
• Steps to run
• Future Improvements
• Contributions
• License
• Contact

🎯 Project Details:

Aim:

To provide an interactive chatbot interface that leverages the power of large language models (LLMs), allowing seamless provider switching and real-time response handling.
It also serves as a way to apply my learnings from the Langchain ecosystem into a practical and useful project.

Features:

• 🔄 Switch between multiple LLM providers:

  • OpenAI
  • Groq
  • Google
  • Ollama (Local LLM)

• 🧠 Dynamically lists available models based on the selected provider.

• 🔎 Supports LangSmith tracing for better observability and debugging.

• 🛠️ Built using LangChain tools:

  • ChatPromptTemplate
  • ChatMessageHistory + History Trimmer
  • LangChain Expression Language (LCEL)
  • RunnablePassthrough, RunnableWithMessageHistory
  • StringOutputParser

• 🚧 Change LLM provider and model on the fly with zero restarts.

Screenshot:

• Choosing a provider:

• Dynamic model listing:

• Chatbot interface:

Tech Stack:

• Frontend: Streamlit
• LLM Orchestration: LangChain

🚀 Steps to run:

1. Clone the repository:

   git clone --depth 1 https://github.com/Bbs1412/QnA_ChatBot

2. Set up virtual environment:

   cd QnA_ChatBot
   python -m venv venv
   
   venv\Scripts\activate
   # or
   source venv/bin/activate
   
   pip install -r requirements.txt

3. Set the environment variables (Optional):

   • Either create a .env file in the root directory:

     OPENAI_API_KEY=your_openai_api_key
     GROQ_API_KEY=your_groq_api_key
     GEMINI_API_KEY=your_gemini_api_key

   • Or use Streamlit's config (.streamlit/secrets.toml):

     [OpenAI]
     API_KEY = "your_openai_api_key"
     
     [Groq]
     API_KEY = "your_groq_api_key"
     
     [Google]
     API_KEY = "your_google_api_key"

4. Run the app:

   streamlit run app.py

📈 Future Improvements:

• Add sliders to let users control temperature, max output tokens and context size dynamically.
• Ensure these parameters are only passed to the LLM when explicitly set by the user.
• This avoids overriding the default values defined by some providers.
• (e.g., certain Ollama models), which differ from standard defaults like 1.
• Add Google provider.

🤝 Contributions:

Any contributions or suggestions are welcome!

📜 License:

• This project is licensed under the MIT License.
• See the LICENSE file for details.

📧 Contact:

• Email - bhushanbsongire@gmail.com