 So, where are we overall trying to get to? We want to implement a deep learning-based game-claying system that, while approximates alpha 0, and to build such a deep learning system, we will be using PyTorch. Now, we will do a super-fast introduction to PyTorch in just 5 minutes. At a very high level, PyTorch is just like NumPy, with a focus on tensors and gradients. And we will now do a few simple things with tensors. Bear with me here, we will do a lot on PyTorch in week 2, and you will feel much more comfortable about it in the second week.