Figuring what data augmentation split we would be using

[beffiong1]

On getting the data we need to decide what form of training/test split we would be doing

[Manaswini1208]

I feel that, a common practice is to allocate 70% or 80% of the data for training and the remaining 30% or 20% for testing. This split is often used when you have a moderate-sized dataset. In some cases, where the dataset is relatively large, we might choose a 60% training and 40% testing split. This allows for more data to be used for training, which can be beneficial if our model is complex and requires a larger amount of training data.