Missing requirements/imports

[rhsimplex]

Hi, thank you for the great paper and sharing your code.

Sorry I don't have time to make a small PR, but there are a couple requirements and imports missing that you'll notice if you try to run the train_attribute_and_evaluate notebook. If create your environment with conda env create, you will not be able to run the notebook because you're missing the ml-collections and more-itertools libraries.

Similarly, you'll find the time library has not been imported, and everywhere you use tqdm as module instead of the right function (tqdm vs tqdm.tqdm).

Thanks again,
Ryan

[josejimenezluna]

Along the lines of what @rhsimplex has pointed out, specific version numbers (also for python) are missing in the environment file.

This causes conda env create -f environment.yml to take forever in solving the right environment.