PyTorch Advanced — all 44 problems

Advanced PyTorch techniques for production and research. Master custom autograd functions, gradient manipulation, hooks, model serialization, einsum operations, weight initialization, learning rate schedulers, and custom optimizers.