Drain DQN on Inverted Domain

oliverc1623 / DRIVE-Sim

A PyTorch-based framework to conduct deep reinforcement learning research in multiple autonomous vehicle simulators

2 stars 1 forks source link

Open oliverc1623 opened 3 months ago

oliverc1623 commented 3 months ago

Work with this file: https://github.com/oliverc1623/DRIVE-Sim/blob/main/ser/minigrid/cuda-ddqn-minigrid.py First try to modify lines 122-130 (the preprocess function) so that the model learns on inverted images. To invert 255-rgb image.