Apologies for the state of this codebase.

Pre-trained models (incl. the random forests) are available in the "models-2023-04-10-2229.pkl" file. You'll need cuML to run them. You'll find an example of how to use them in the "clsp" function in 02-modelling.py.

Basically: you'll need the msmarco-distilbert-base-v4 sentence transformer, the models, and some sentences.