Multi-Session Chatbot

A simple AI chatbot built using LangChain and Ollama (qwen2.5:0.5b) with support for multiple independent chat sessions.

Features

• Chat with qwen2.5:0.5b locally using Ollama
• Maintains conversation history
• Create multiple chat sessions
• Switch between sessions
• View all available sessions
• View the current session
• Each session has its own independent memory

Technologies Used

• Python
• LangChain
• Ollama
• qwen2.5:0.5b

How It Works

The chatbot uses RunnableWithMessageHistory to manage conversation history.

Each session has a unique session_id, and the conversation history is stored separately for each session.

User
  ↓
Chatbot
  ↓
Session ID
  ↓
Chat History
  ↓
qwen2.5:0.5b
  ↓
Response

Session Commands

Command	Description
1	Create a new session
2	Show all sessions
3	Show current session
4	Switch session
break	Exit chatbot

Example

You: my name is Vignesh
Bot: Nice to meet you, Vignesh!

1
Enter a new session_id: user2

You: my name is Kumar
Bot: Nice to meet you, Kumar!

4
Give the session_id to change: user1

You: what is my name?
Bot: Your name is Vignesh.

Each session maintains its own conversation history.

Installation

Install the required packages:

pip install langchain langchain-core langchain-ollama

Make sure Ollama is installed and the Llama 3.1 model is available:

ollama pull qwen2.5:0.5b

Run the Python program:

python chatbot.py

Project Structure

chatbot/
│
├── chatbot.py
└── README.md

Learning Concepts

This project demonstrates:

• LangChain prompts
• Chat message history
• Session management
• RunnableWithMessageHistory
• Local LLM integration
• Ollama
• Conversation memory