DeepSeek on Your Local Machine (Web Version)

Project Overview

This project consists of a backend and a frontend that enable interaction with the DeepSeek AI model locally.

By using Docker and the Ollama service, AI processing can be done without relying on cloud services, providing a faster and more secure experience.

Key Features:

• Backend: Developed with .NET 9, handling chat requests and fetching responses from DeepSeek.

• Frontend: Built with React, providing a modern and user-friendly interface for interaction.

• Docker Compose: Simplifies service management and enables easy project execution.

• Optimized Models: Supports DeepSeek and other available versions.

How to Use

1. Install and Run Ollama

First, install Ollama on your system. Then, to download and run the lightweight version of DeepSeek, execute the following commands:

ollama run deepseek-r1:1.5b

You can use other version :

ollama run deepseek-r1:671b

These commands will download and run the smallest DeepSeek model, making it suitable for efficient processing with minimal resource consumption.

2. Run the Project with Docker Compose

To start the backend and frontend services, run the following command:

docker-compose up -d

This command will launch both services, allowing you to run the project without additional setup.

3. Access the Project

Once running, you can access the web interface by navigating to:

http://localhost:3002

Project Benefits

✅ Run DeepSeek locally without relying on cloud services, ensuring better security and speed.

✅ UI similar to popular chatbots for a seamless user experience.

✅ Docker-based architecture for easy deployment and scalability.

✅ Supports optimized AI models to reduce hardware resource consumption.

✅ Easily extendable to support additional AI models in the future.

✅ Fast message processing and real-time response to user queries.