Once the SyntheticRegressionData class is defined, we can instantiate it with predefined ground truth parameters to create our synthetic dataset. For example, by specifying the true weight vector as $\mathbf{w} = [2, -3.4]^ op$ and the bias as $b = 4.2$, the resulting data object will store the generated features and labels. Inspecting an individual data point from this instantiated object, such as the first entry, will return a feature vector in $\mathbb{R}^2$ and its corresponding computed scalar label.