How to Install DeepSeek on Your Cloud Server with Ollama LLM

Running large language models like DeepSeek on your cloud server or VPS is a powerful way to explore AI capabilities while maintaining control over your infrastructure. This guide will walk you through installing DeepSeek using Ollama on a cloud-hosted Ubuntu 24.04 instance, setting up automated deployments with DeployHQ, and configuring a Web UI for remote access.

What is DeepSeek, Ollama, and DeployHQ?

• DeepSeek: An advanced AI model designed for natural language processing tasks like answering questions, generating text, and more.

• Ollama: A platform that simplifies running large language models on your server by providing tools to manage and interact with models like DeepSeek. More here.

• DeployHQ: A deployment automation platform that integrates with Git repositories to manage and automate deployments across different environments.

• Web UI: A graphical interface that allows you to interact with DeepSeek through your browser from anywhere.

Prerequisites

Before we begin, make sure you have:

• A VPS or cloud instance running Ubuntu 24.04
• Root or sudo access to your server
• At least 16GB of RAM (32GB recommended for production use)
• Domain name (optional but recommended)
• Git account
• DeployHQ account
• Basic understanding of server administration

Step 1: Initial Server Setup

First, connect to your VPS/cloud server via SSH:

ssh username@your-server-ip

Update your system:

sudo apt update && sudo apt upgrade -y

Install essential packages:

sudo apt install python3 python3-pip git ufw

Configure basic firewall:

sudo ufw allow OpenSSH
sudo ufw allow 80/tcp
sudo ufw allow 443/tcp
sudo ufw allow 8080/tcp  # For Web UI if not using reverse proxy
sudo ufw enable

Step 2: Configure Domain and SSL (Optional but Recommended)

If you're using a domain name:

# Install Nginx
sudo apt install nginx

# Install Certbot for SSL
sudo apt install certbot python3-certbot-nginx

# Configure SSL
sudo certbot --nginx -d yourdomain.com

Step 3: Install Ollama for DeepSeek

Install Ollama:

curl -fsSL https://ollama.com/install.sh | sh
ollama --version

Configure system service:

sudo systemctl start ollama
sudo systemctl enable ollama

Step 4: Set Up Version Control and Automated Deployment

1- Create a deployment user:

sudo adduser deploy
sudo usermod -aG sudo deploy

2- Configure SSH keys for DeployHQ:

# As deploy user
su - deploy
mkdir -p ~/.ssh
touch ~/.ssh/authorized_keys
chmod 700 ~/.ssh
chmod 600 ~/.ssh/authorized_keys

3- Set up project structure:

mkdir -p /var/www/deepseek
chown deploy:deploy /var/www/deepseek

4- Initialise Git repository:

cd /var/www/deepseek
git init

5- Create deployment configuration:

deployment:
  environments:
    production:
      branch: main
      server: production-server
    # Only if we need a staging server, for example to test new models, etc
    staging:
      branch: staging
      server: staging-server

  commands:
    pre_deploy:
      - python3 -m venv venv
      - source venv/bin/activate
      - pip install -r requirements.txt

    post_deploy:
      - sudo systemctl restart ollama
      - sudo systemctl restart nginx

Step 5: Download and Configure DeepSeek Model

# Here you could download and configure other models as well such as: Mistral, Llama, etc
ollama run deepseek-r1:7b

Create model configuration:

# config/model_config.yml
model:
  name: deepseek-r1
  version: 7b
  parameters:
    temperature: 0.7
    max_tokens: 2048

Step 6: Set Up Web UI with Nginx Reverse Proxy

Install Web UI dependencies:

python3 -m venv venv
source venv/bin/activate
pip install open-webui

Configure Nginx reverse proxy:

# /etc/nginx/sites-available/deepseek
server {
    listen 80;
    server_name yourdomain.com;

    location / {
        proxy_pass http://localhost:8080;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
    }
}

Enable the site:

sudo ln -s /etc/nginx/sites-available/deepseek /etc/nginx/sites-enabled/
sudo nginx -t
sudo systemctl restart nginx

Step 7: Continuous Deployment Workflow

1- In DeployHQ dashboard:

• Create new project
• Connect Git repository
• Add server credentials
• Configure deployment paths

2- Setup automatic deployments: It can be done directly on DeployHQ dashboard, in which a webhook to the repository needs to be installed.

Monitoring and Maintenance

1- Set up monitoring:

# Install monitoring tools
sudo apt install htop prometheus node-exporter

# Configure system monitoring
sudo systemctl enable prometheus node-exporter

2- Create backup script:

#!/bin/bash
# backup.sh
backup_dir="/backup/deepseek"
date=$(date +%Y%m%d)
tar -czf "$backup_dir/model_data_$date.tar.gz" /var/www/deepseek/data

Security Considerations

1- Configure rate limiting in Nginx:

limit_req_zone $binary_remote_addr zone=one:10m rate=1r/s;

2- Set up fail2ban:

sudo apt install fail2ban
sudo systemctl enable fail2ban

Conclusion

You now have a production-ready DeepSeek installation running on your cloud server with automated deployments through DeployHQ. The setup includes:

• Secure access through SSL
• Automated deployments
• Monitoring and backup solutions
• Rate limiting and security measures

Remember to regularly update your system and monitor resource usage, especially when running resource-intensive language models in a production environment.

Happy deploying!