The Training Set is the set given to the model for training, analyzing, and learning. The Test Set is the set that is used for testing the model locally before using it in a real application. The Training Set is Labeled data, and the Test Set has no labels.
It is important to divide the dataset into training and test sets so that the model will not go in overfitting or underfitting conditions. Also, it is a great method to evaluate the model and understand the characteristics of the data. In most cases, the split is 70/30, meaning 70% of the full data set for training and 30% for testing.