flask_ollama_web

This Python module spanns a web server up that interacts with a AI system running locally and listening at http://localhost:11434/api/chat. There the questions will be asked.

Install

You need a local ollama installation and a 'running' ollama instance to connect to.

Once you have installed ollama you can download and start whichever model ollama supports. E.g. start with a rather small model like llama3

ollama run llama3

With this you can install the flask ollama frontend like this:

pip install git+https://github.com/stela2502/flask_ollama_web.git

Usage

echo "from flask_ollama_web.routes import app


if __name__ == \"__main__\":
    app.run(host=\"0.0.0.0\", port=8080)" > /app/run.py

And then simply

 /app/run.py

 * Running on all addresses (0.0.0.0)
 * Running on http://127.0.0.1:8080
 * Running on http://<external IP>:8080

This will start a single flask server that you can connect to. And interact in a very simplye way:

A simple oolama web interface

One big benefit from this is that you can see the answer in html and copy it in markdown ;-)