Conv Parameter Count

Problem

Count the trainable parameters of a 2D convolution layer.

The weight tensor has cin * cout * k * k entries; when bias is True add cout bias values. Return the integer total.

case 1

in: [3,16,3,true]

expected: 448

case 2

in: [1,1,3,false]

expected: 9

case 3

in: [3,64,3,true]

expected: 1792

case 4

in: [64,128,1,false]

expected: 8192

Starter codePython

def conv_param_count(cin, cout, k, bias):
    # Your code here
    pass

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