def stochasticGradientDescent(net: network.NeuralNet, inputs, targets,
epochs: int):
assert (len(inputs) == len(targets))
indices = list(range(0, len(inputs)))
for i in range(0, epochs):
print("Epoch %d" % i)
random.shuffle(indices)
for i in indices:
grads = net.gradient(inputs[i], targets[i])
for grad in grads:
if grad is not None:
grad *= 0.1
net.adjustWeights(grads)
