def cross_entropy(pred, label, axis=1):
n0 = pred.ndim
n1 = label.ndim
assert n0 == n1 + 1, (
"target ndim must be one less than input ndim; input_ndim={} "
"target_ndim={}".format(n0, n1))
num_classes = pred.shapeof(axis)
# Denominator of the softmax
#offset = zero_grad(pred.max(axis=axis, keepdims=True))
#pred = pred - offset
#down = mgb.opr.elem.exp(pred).sum(axis=axis, keepdims=True)
up = indexing_one_hot(pred, label, axis, keepdims=True)
return (-log(up + 1e-8)).mean()
def cross_entropy_with_softmax(pred, label, axis=1, mask=None):
n0 = pred.ndim
n1 = label.ndim
assert n0 == n1 + 1, (
"target ndim must be one less than input ndim; input_ndim={} "
"target_ndim={}".format(n0, n1))
num_classes = pred.shapeof(axis)
# Denominator of the softmax
offset = zero_grad(pred.max(axis=axis, keepdims=True))
pred = pred - offset
down = mgb.opr.elem.exp(pred).sum(axis=axis, keepdims=True)
up = indexing_one_hot(pred, label, axis, keepdims=True)
if mask is None:
return (log(down) - up).mean()
else:
return ((log(down) - up) * mask).sum() / mask.sum()