Issues Found in "Prepare Data for Modeling" Section of "Train deep learning models in Azure Databricks""

Question

The "Prepare data for modeling" section of the page Train deep learning models in Azure Databricks appears to have two issues:

1. The use of td.TensorDataset(train_x, train_y) where the "td" alias isn't defined anywhere in the content.
2. A discrepancy in the statement "The loaders in this example split the data into batches of 30, which are passed to the forward function during training or inference," as the example actually uses a batch size of 20.

This question is related to the following Learning Module