Search⌘ K
AI Features

Data Preparation

Understand how to prepare data by splitting datasets into training, validation, and test subsets to effectively evaluate and train image classification models using PyTorch.

Spliting datasets

After selecting a public dataset or creating our own custom one, we have to split the dataset first into three different subset categories, which are the following:

  • Train dataset
  • Validation dataset
  • Test dataset

Train dataset

This dataset is used during training by showing data and its labels to the model. At the end of each epoch, this dataset is also used to calculate the accuracy and loss by not updating the ...