ChatBot Project

Overview

This project is an AI-powered chatbot that uses the langchain4j library to interact with the local Ollama API. The chatbot is designed to provide intelligent responses and maintain conversation history using in-memory storage.

Installation and Launching Instructions

Prerequisites

• Docker
• Java 21
• Maven

Step 1: Launch Docker Compose

To start the Ollama and qdrant service using Docker Compose, navigate to the project directory and run the following command:

docker compose up -d

Step 2: Pull the LLMs

To pull the mistral and codellama model, use the following command:

docker exec -it ollama ollama pull mistral
docker exec -it ollama ollama pull codellama

Step 3: Initialize Qdrant (required for RAG/AST cases only)

To set up qdrant please use either the QdrantStorageMarkdownInitializer for indexing a markdown file or the QdrantStorageJavaInitializer for indexing a Java code base. In both cases you need to edit the class file by setting the fully qualified path to the respective folder and execute the main method.

Step 4a: Launch the Chatbot (Basic ChatBot only)

To launch the chatbot, follow these steps:

1. Navigate to the project directory.

2. Build the project using Maven:

   mvn clean install

3. Run the chatbot application:

   java -jar target/chatBot-1.0-SNAPSHOT.jar

This will start the chatbot, and you can interact with it via the console. Type 'exit' to quit the application.

Step 4b: Run the Chatbot in the IDE (Basic and RAG/AST cases)

To run the chatbot in IntelliJ IDEA, follow these steps:

1. Open IntelliJ IDEA.
2. Open the project by selecting File > Open and navigating to the project directory.
3. Wait for IntelliJ IDEA to index the project and download dependencies.
4. Navigate to the *ChatBot file.
5. Run its main method.

This will start the chatbot, and you can interact with it via the console in the IDE. Type 'exit' to quit the application.