def test_dsoftmax_matrix():
tensor_shape = [random.randint(2, 10) for _ in range(2)]
helpers.check_gradient(lambda X: calculate.softmax(X),
lambda X: calculate.dsoftmax(calculate.softmax(X)),
f_arg_tensor=numpy.random.random(tensor_shape),
f_shape='jac-stack')
def derivative(self, input_vec, output_vec):
"""Return the derivative of this function.
We take both input and output vector because
some derivatives can be more efficiently calculated from
the output of this function.
"""
return calculate.dsoftmax(output_vec)
def test_dsoftmax():
helpers.check_gradient(calculate.softmax,
lambda x: calculate.dsoftmax(calculate.softmax(x)),
f_shape='jac')