Implementation of Multilayer Perceptron from Scratch¶

In [1]:

import d2l
from mxnet import np, npx, gluon
npx.set_np()

batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

Initialize model parameters

In [2]:

num_inputs, num_outputs, num_hiddens = 784, 10, 256

W1 = np.random.normal(scale=0.01, size=(num_inputs, num_hiddens))
b1 = np.zeros(num_hiddens)
W2 = np.random.normal(scale=0.01, size=(num_hiddens, num_outputs))
b2 = np.zeros(num_outputs)
params = [W1, b1, W2, b2]

for param in params:
    param.attach_grad()

Activation Function

In [3]:

def relu(X):
    return np.maximum(X, 0)