Skip to content

Prework

Please complete the following setup steps before the workshop.


Local Models

Install Ollama

Note

To run the Granite model locally, we recommend having at least 16GB of RAM for optimal performance.

To run models locally on your machine:

  1. Download and install Ollama: https://ollama.com/download

  2. Pull the models

    Please pull the models before the workshop!

    ollama pull ibm/granite4:micro-h
    ollama pull ibm/granite-embedding:30m
    ollama pull ibm/granite-docling
    
  3. Quick test

    Test by chatting with a model (e.g., ask what model it is):

    ollama run granite4:micro "what model are you and who created you?"
    

    Note

    Don't be surprised by hallucination and different answers each time you run this.


Python Environment Manager

Install uv

We will be using uv as your Python package and environment manager. If you’re unfamiliar with uv, refer to the uv installation guide.

  • uv is a fast and modern alternative to pip and virtualenv, fully compatible with both
  • uv manages versioned installations of Python

For most people, the install guide provides a one-line install command:

On macOS/Linux

curl -LsSf https://astral.sh/uv/install.sh | sh

On Windows

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Get the workshop code

Option A: Clone with Git (recommended):

git clone https://github.com/IBM/beeai-workshop.git

Option B: Download ZIP:

If you're not comfortable with Git, download the ZIP file and extract it to your desired location.


Install project Python dependencies

Preinstall all the packages used in the demo, installed in the correct virtual environment using uv.

  1. Navigate to the specific demo folder for this workshop:

    cd beeai-workshop/opentech
    
  2. Install all required python dependencies for the 4 sub-projects:

    uv sync --directory docling
    uv sync --directory mellea
    uv sync --directory beeaiframework
    uv sync --directory agentstack
    

Install Open WebUI

Once uv is installed, use uvx to run Open WebUI with Python 3.11 (recommend for Open WebUI).

macOS/Linux

DATA_DIR=~/.open-webui uvx --python 3.11.11 open-webui@0.6.36 serve

Windows

$env:DATA_DIR="C:\open-webui\data"; uvx --python 3.11.11 open-webui@0.6.36 serve

Once the downloads, install, and start are complete, you will have a fancy Open WebUI "get started" page at localhost:8080. You can now kill the server with control-C back in the terminal. We'll do the setup during the workshop.


Install Agent Stack

Ideally, everyone will get hands-on experience wrapping all the pieces together with Agent Stack. If you cannot run Agent Stack on your laptop, come to the workshop anyway! We will demonstrate.