1. Learn
    2. /
  2. Courses
    3. /
  3. Introduction to Deep Learning with PyTorch

Exercise

Using the MSELoss

Recall that we can't use cross-entropy loss for regression problems. The mean squared error loss (MSELoss) is a common loss function for regression problems. In this exercise, you will practice calculating and observing the loss using NumPy as well as its PyTorch implementation.

The torch package has been imported as well as numpy as np and torch.nn as nn.

Instructions

100 XP
  • Calculate the MSELoss using NumPy.
  • Create a MSELoss function using PyTorch.
  • Convert y_hat and y to tensors and then float data types, and then use them to calculate MSELoss using PyTorch as mse_pytorch.