In this blog post, we learn how neural networks work by building them from scratch. We also implement loss functions and train them by deriving gradients without using any libraries. We will use raw PyTorch to accomplish this task. Check out the blog to know more.