A clean implementation of the Deep Deterministic Policy Gradient Algorithm using PyTorch. Referenced from https://github.com/philtabor/Youtube-Code-Repository/tree/master/ReinforcementLearning/PolicyGradient/DDPG/pytorch/lunar-lander
Fixed inconsistencies with the original paper. Refactored the code and added epsilon-greedy exploration strategy, which improved training performance
