#meta-learning

A PyTorch tutorial demonstrating how to build a memory-augmented agent that uses differentiable memory, prioritized experience replay and meta-learning to adapt continuously without catastrophic forgetting.