Set up a Local AI like ChatGPT on your own machine!

[thumbtak]

Install on Linux

$ sudo apt install python3.11-venv

$ python3.11 -m venv myenv

$ source myenv/bin/activate

$ pip install open-webui

$ pip install async_timeout

$ sudo docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main

$ ollama pull llama3.1:latest

$ open-webui serve

http://localhost:8080

Run on Linux, (after the above has been closed):

$ cd ~/myenv

$ source myenv/bin/activate

$ open-webui serve

http://localhost:8080

Allow other on your network to use this:

http://your-ip:8080

Script that can be run each time you want to start the Web UI:

#!/bin/sh

# Navigate to the home directory, or the directory where your virtual environment is
cd ~/myenv || exit # If this fails, exit the script

# Activate the virtual environment
if [ -f "bin/activate" ]; then
. bin/activate # Use '.' (dot) instead of 'source' in sh
else
echo "Virtual environment not found. Exiting."
exit 1
fi

# Run open-webui
open-webui serve

# Notify user the server is up
echo "Activated, open at http://localhost:8080"

Save as a sh file and run with $ sh file.sh command.

More Information:

https://github.com/open-webui/open-webui

https://ollama.com/

• This topic was modified 2 months, 4 weeks ago by thumbtak. Reason: Removed sudo from the pip commands as "Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager."
• This topic was modified 2 months, 3 weeks ago by thumbtak. Reason: I removed s command by putting a strickthrough
• This topic was modified 2 months, 3 weeks ago by thumbtak. Reason: Removed repeated command

[thumbtak]

Upgrade Open-WebUI:

Start Environment:

$ python3.11 -m venv myenv

$ source myenv/bin/activate

Update Pip:
$ pip install --upgrade pip

Update Open WebUI:
$ pip install --upgrade open-webui

Restart the Server:
$ open-webui serve

Note:  Make sure it is not already running.

• This reply was modified 2 months, 3 weeks ago by thumbtak. Reason: Updated the command, to working. Update should now work
• This reply was modified 2 months, 2 weeks ago by thumbtak. Reason: Remove command replaced with uninstall command
• This reply was modified 3 weeks, 5 days ago by thumbtak. Reason: Update command fixed

[thumbtak]

Retraining vs RAV vs Contract: Your Local Data on LLMs!

[thumbtak]

The commands were updates as the old ones didn’t work.

[thumbtak]

If you want to run Ollama in a terminal window, use the following command.

$ ollama run model

Example:
$ ollama run mistral

[Author]

Posts