Welcome to Jetson Orin NANO Developer Kit repository
02.Install large model dialogue platform
Open WebUI
Open WebUI1. Environmental Requirements2. Docker Setup2.1. Official Docker Installation2.2. Installing Docker from Alibaba Cloud2.3. Adding Access Permissions3. Open WebUI Installation4. Running Open WebUI4.1. Administrator Account4.2. Registering and Logging In4.3. User Interface5. Model Conversation5.1. Switching Models5.2. Demonstration: LLaVA6. Common Problems6.1. Closing Open WebUI6.2. Common ErrorsUnable to Start Open WebUIService Connection Timeout
Open WebUI is an open source project that aims to provide a simple and easy-to-use user interface (UI) for managing and monitoring open source software and services.
When using Open WebUI, there is a high probability that the conversation will be unresponsive or timeout. You can try restarting Open WebUI or switching to the Ollama tool to run the model!
1. Environmental Requirements
To install Open WebUI on the host with Conda, you need Node.js >= 20.10 and Python 3.11:
| Environment setup method | Difficulty (relatively) |
|---|---|
| Host | High |
| Conda | Medium |
| Docker | Low |
Note: Open WebUI is already set up on the factory system. If you need to set it up yourself, you can refer to the steps for installing Open WebUI through Docker below.
2. Docker Setup
2.1. Official Docker Installation
If Docker is not installed, you can use a script to install Docker with one click.
xxxxxxxxxxUsers in China may not be able to install Docker in the official way. We recommend installing Docker from Alibaba Cloud or using our image directly.
Update the local package list
xxxxxxxxxxsudo apt update
Upgrade the installed packages
xxxxxxxxxxsudo apt upgrade
Download and run the script
Download the get-docker.sh file and save it in the current directory.
xxxxxxxxxxsudo apt install curl
xxxxxxxxxxcurl -fsSL https://get.docker.com -o get-docker.sh
Run the get-docker.sh script file with sudo privileges.
xxxxxxxxxxsudo sh get-docker.sh
2.2. Installing Docker from Alibaba Cloud
If you cannot install it yourself, please use the image we provide.
Update the local package list
xxxxxxxxxxsudo apt update
Install the required software
xxxxxxxxxxsudo apt install apt-transport-https ca-certificates curl gnupg2 lsb-release software-properties-common
Add the GPG key of the software source
xxxxxxxxxxcurl -fsSL https://mirrors.aliyun.com/docker-ce/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg
Add the Alibaba Cloud mirror software source
xxxxxxxxxxecho "deb [arch=arm64 signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://mirrors.aliyun.com/docker-ce/linux/ubuntu noble stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
Install Docker
xxxxxxxxxxsudo apt updatesudo apt install docker-ce docker-ce-cli containerd.io docker-compose-plugin
2.3. Adding Access Permissions
Add access permissions for the current system user to the Docker daemon: the Docker command can be used without the sudo command
xxxxxxxxxxsudo usermod -aG docker $USERnewgrp docker
3. Open WebUI Installation
For a system with Docker installed, you can directly enter the following command in the terminal: the image shows the result after pulling
xxxxxxxxxxdocker pull ghcr.io/open-webui/open-webui:main

4. Running Open WebUI
The factory setting is to start it automatically at boot, so you can access it directly with a browser. If you finish the deployment inside your own system, enter the following command in the terminal to start the specified Docker:
xxxxxxxxxxdocker run --network=host -v open-webui:/app/backend/data -e OLLAMA_BASE_URL=http://127.0.0.1:11434 --name open-webui --restart always ghcr.io/open-webui/open-webui:main

After a successful startup, access it in the browser with the following URL:
xxxxxxxxxxhttp://localhost:8080/
The same LAN can use the mainboard IP:8080 to access:
xxxxxxxxxxAssume the mainboard IP is 192.168.2.105, we can access it through 192.168.2.105:8080
4.1. Administrator Account
For first-time use you need to register an account yourself. This account is an administrator account, and the information can be filled in as required!
xxxxxxxxxxSince all the contents of our image have been set up and tested, users can directly log in with our registered account:Username: yahboomEmail: yahboom@163.comPassword: yahboom
4.2. Registering and Logging In



4.3. User Interface

5. Model Conversation
Using Open WebUI for conversation is slower than directly running the model with the Ollama tool, and may even cause service connection timeout failures. This is related to the memory of the Jetson mainboard and cannot be avoided!
xxxxxxxxxxUsers who have ideas can switch to another Linux environment to set up the Ollama tool and the Open WebUI tool for conversation
5.1. Switching Models
Click Select a model to select the specified model for conversation.
xxxxxxxxxxThe model pulled by ollama will be automatically added to the Open WebUI model options. Refresh the web page and the new model will appear!

5.2. Demonstration: LLaVA
The LLaVA case demonstrated requires 8G or even more than 8G to run. Users can use other cases to test the Open WebUI conversation function!
xxxxxxxxxxWhat's in this image?

6. Common Problems
6.1. Closing Open WebUI
Close the automatically started Open WebUI.
View the running Docker
xxxxxxxxxxdocker ps
Close the running Docker
xxxxxxxxxxdocker stop [CONTAINER ID] # Example docker stop 5f42ee9cf784
View the stopped container
xxxxxxxxxxdocker ps -a
Clean up the stopped container
xxxxxxxxxxdocker rm [CONTAINER ID] # Example docker rm 5f42ee9cf784
Clean up all stopped containers:
xxxxxxxxxxdocker container prune
6.2. Common Errors
Unable to Start Open WebUI
docker: Error response from daemon: Conflict. The container name "/open-webui" is already in use by container "cfc05c84f8e38b290337e7178c76fd1c49076f94b11ed3d49d9448be72b7f20f". You have to remove (or rename) that container to be able to reuse that name.
Solution: close Open WebUI once and restart it!
Service Connection Timeout
Open WebUI: Server Connection Error
Close Open WebUI once and restart it, then ask again or switch to the Ollama tool to run the model and ask questions!