Setup and Reproducibility Instructions

0. Python

Make sure you have python 3.11 installed.

1. Install dependencies with uv

First, ensure you have uv installed. You can install it via pip:

pip install uv

Then, install all dependencies specified in pyproject.toml and uv.lock:

uv pip install -r pyproject.toml

1.5

You might need to check the PyTorch compatibility with your CUDA version. Information available here: https://pytorch.org/get-started/

2. Launch Jupyter and run experiments

Start Jupyter Lab or Notebook:

uv run -- jupyter lab

or

uv run -- jupyter notebook

or You can also activate the virtualenv manually:

source .venv/bin/activate

and then run jupyter

Open experiments.ipynb and run all cells to reproduce the results.

---

Note: All dependencies are pinned in uv.lock, versions can be found in pyproject.toml.