This is a collection of examples and tutorials for using the RL4CO library.

The root directory is made of quickstarts and contains the following:

This is the root directory of the examples. The following quickstarts are available:

• 1-quickstart.ipynb: here we train a model on a simple environment - it takes less than 2 minutes!

• 2-full-training.ipynb: similar to the previous notebooks but with a more interesting environment, with checkpointing, logging, and callbacks.

  • 2b-train-simple.py: here we show a simple script that can be called with python 2b-train-simple.py. This is simplified and does not use Hydra - for those who prefer a simpler setup. Note that we also made a Hydra tutorial here.

• 3-creating-new-env-model.ipynb: here we show how to extend RL4CO to solve new problems and create new models from zero to hero!

Under the modeling/ directory, here are some additional examples for modeling and inference.

Under the datasets/ directory, here are some additional examples for using your custom data to train/evaluate your models

Under the advanced/ directory, here are some additional examples for advanced topics.