How to Set Up Ollama and Open WebUI for Remote Access - Your Personal Assistant on the Go!

---

---

AI Nuggets

Instructions for Running Large Language Models Locally and Accessing Them Remotely

Requirements

• olama: A framework for running large language models locally.
• Docker: Software for creating containers to run applications.
• open web UI: An open-source tool for interacting with large language models via a web interface.
• ngrok: A tool that provides a public URL to local web applications.

Steps

1. Install olama
   • Download olama by clicking the download button on their website.
   • Verify installation on a Mac by checking for the olama icon in the menu bar.
2. Download Models for olama
   • Download models from the olama website.
   • Deploy models using the terminal command: olama run <model_name>.
   • For detailed instructions, refer to the dedicated video linked in the video description.
3. Install Docker
   • Visit the Docker website and download the application for your operating system.
   • Complete the setup process and accept the user agreement.
   • Verify Docker installation by looking for the Docker icon on the menu bar.
4. Set Up open web UI
   • Go to the open web UI repository on GitHub (link in the video description).
   • Copy the installation command from the installation section.
   • Paste the command into the terminal to install open web UI in Docker.
   • Confirm installation by navigating to http://localhost:3000 in a web browser.
   • Sign up for an account with credentials that remain local to your machine.
   • For more guidance, refer to the video tutorial linked in the description.
5. Install ngrok
   • Sign up for an account on the official ngrok website (ngrok.com).
   • Follow the installation instructions for your operating system.
   • If using Mac OS, use the Homebrew command provided on the ngrok page.
   • Authenticate your local machine using the authentication code from the ngrok page.
   • Install and set up ngrok on your machine.
6. Create a Public URL with ngrok
   • Copy the command from the ngrok installation page to expose your local URL publicly.
   • Change the port number in the command to match the port open web UI is running on (e.g., 3000).
   • Paste the command into the terminal.
   • Copy the forwarding URL generated by ngrok.
   • Confirm the URL works by entering it in a new browser window.
7. Access open web UI Remotely
   • Share the ngrok forwarding URL with your phone and open it in a browser.
   • Log in with your user credentials.
   • Upload a photo and use a model like “lava” to analyze and describe the image.
8. End ngrok Session
   • To stop the URL forwarding, go back to your terminal and press Ctrl + C.

Additional Notes

• To keep the system running, ensure olama, open web UI, and ngrok are active on your machine.
• The entire process is accessible remotely without requiring installation on your phone.

Video Resources

• Dedicated video explaining olama setup: Linked in the video description.
• Video tutorial for open web UI installation: Linked in the description.

URLs and Commands

• olama download: Not provided, refer to the olama website.
• Docker download: Not provided, refer to the Docker website.
• open web UI GitHub repository: Not provided, refer to the video description for the link.
• ngrok sign up and installation: Visit ngrok.com and follow the instructions provided.

Tips

• Keep your ngrok authentication code confidential.
• Change the port number in the ngrok command to match the service you are forwarding.
• Use Ctrl + C to stop ngrok when you’re done.