UiPath AI Specialist v1 — Question 171

When creating a training dataset, what is the recommended number of samples for a Column field?

Answer options

Correct answer: C

Explanation

The correct answer is C, as having 50-200 document samples per column field is generally considered sufficient to ensure diversity and representation in the dataset, which is crucial for effective training. Options A and B provide too few samples to capture the necessary variability, while option D suggests an excessive number that may not be necessary for effective training.