Set up environment

We recommend you to set up a conda environment for packages used in this homework.

For part II, you need access to GPUs.

You can either use HPC server (see documentation here[https://github.com/Athul-R/NYU-HPC-Tutorials]) or use API services like lightning.ai. If using such service, you can skip conda commands and just install requirements in your studio.

conda create -n 2590-hw3 python=3.12.11
conda activate 2590-hw3
pip install -r requirements.txt

To register a jupyter kernel from your conda environment, use

conda activate 2590-hw3
python3 -m ipykernel install --user --name envkernel