Batch Padding Waste

Problem

Compute wasted padding slots when batching: ceil(n_requests / batch_size) * batch_size - n_requests. Non-positive arguments return 0.

Hint

Padding keeps tensor shapes fixed but wastes compute on empty slots.

case 1

in: [100,32]

expected: 28

case 2

in: [64,8]

expected: 0

case 3

in: [5,7]

expected: 2

case 4

in: [0,4]

expected: 0

Starter codePython

import math

def batch_padding_waste(n_requests, batch_size):
    # Your code here
    pass

4 test cases · run your code against them in the editor