4D-convolution on Python

Question

I don't understand a few things in 4D-convolution. I have input tensor with dimension (5, 3, 17, 5), where:

batch_amount = 5
tensor_Height = 3
tensor_Weight = 17
tensor_Cin = 5

Parameters of kernel (3, 3, 5, 2):

kernel_X = 3
kernel_Y = 3
kernel_depth = 5
kernel_amount or C_out = 2

The bias is (2,):

b = [0.31862102 1.98459104]

How can I implement convolution and forward propagation with the current parameters using only numpy.

I tried to follow the example of this:

def stride_conv(arr1,arr2,s,p):
    beg = 0
    end = arr2.shape[0]
    final = []
    for i in range(0,arr1.shape[0]-1,s):
        k = []
        for j in range(0,arr1.shape[0]-1,s):
            k.append(np.sum(arr1[beg+i : end+i, beg+j:end+j] * (arr2)))
        final.append(k)

    return np.array(final)