Just some experimental code to teach a LSTM network to play rocket league. The bulk of this code is copied from Impossibum's Rocket League Reinforcement Learning Bot Quick Start Tutorial Guide, with some custom rewards applied.
This model is currently in training, but I'll push the checkpoints as it goes.