def dot(left_node, right_node, reduction_axes_count=None, name=None):
# type: (Node, Node, int, str) -> Node
"""Return node which performs generalized dot product of two input nodes.
This operation is capable of performing scalar-tensor, matrix-vector product and matrix
multiplication.
:param left_node: The node providing left hand side data.
:param right_node: The node providing right hand side data.
:param reduction_axes_count: The number of axes to reduce during dot-product.
:param name: The optional name for output node.
:return: The new node performing dot-product on input two nodes.
"""
if reduction_axes_count is None:
return Dot(left_node, right_node)
else:
return Dot(left_node, right_node, reduction_axes_count)
def binary_op(op_str, a, b):
if op_str == '+':
return a + b
elif op_str == 'Add':
return Add(a, b)
elif op_str == '-':
return a - b
elif op_str == 'Sub':
return Subtract(a, b)
elif op_str == '*':
return a * b
elif op_str == 'Mul':
return Multiply(a, b)
elif op_str == '/':
return a / b
elif op_str == 'Div':
return Divide(a, b)
elif op_str == 'Dot':
return Dot(a, b)
elif op_str == 'Equal':
return Equal(a, b)
elif op_str == 'Greater':
return Greater(a, b)
elif op_str == 'GreaterEq':
return GreaterEq(a, b)
elif op_str == 'Less':
return Less(a, b)
elif op_str == 'LessEq':
return LessEq(a, b)
elif op_str == 'Maximum':
return Maximum(a, b)
elif op_str == 'Minimum':
return Minimum(a, b)
elif op_str == 'NotEqual':
return NotEqual(a, b)
elif op_str == 'Power':
return Power(a, b)
def binary_op(op_str, a, b):
if op_str == '+':
return a + b
elif op_str == 'Add':
return ng.add(a, b)
elif op_str == '-':
return a - b
elif op_str == 'Sub':
return ng.subtract(a, b)
elif op_str == '*':
return a * b
elif op_str == 'Mul':
return ng.multiply(a, b)
elif op_str == '/':
return a / b
elif op_str == 'Div':
return ng.divide(a, b)
elif op_str == 'Dot':
return Dot(a, b)
elif op_str == 'Equal':
return ng.equal(a, b)
elif op_str == 'Greater':
return ng.greater(a, b)
elif op_str == 'GreaterEq':
return ng.greater_equal(a, b)
elif op_str == 'Less':
return ng.less(a, b)
elif op_str == 'LessEq':
return ng.less_equal(a, b)
elif op_str == 'Maximum':
return ng.maximum(a, b)
elif op_str == 'Minimum':
return ng.minimum(a, b)
elif op_str == 'NotEqual':
return ng.not_equal(a, b)
elif op_str == 'Power':
return ng.power(a, b)
def dot(left_node, right_node, name=None):
# type: (Node, Node, str) -> Node
"""Return node which performs matrix multiplication of two input nodes."""
return Dot(left_node, right_node)
def relu(op): # type: (Node) -> Node
"""Relu operator."""
return Maximum(op, make_float32_constant_like(0., op))
# Flatten
X1 = Reshape(Input, AxisVector([0, 1, 2]), Shape([bz, 784]))
# Normalize
X2 = X1 / make_float32_constant_like(255., X1)
# Affine 1
W1 = Parameter(float_element_type, Shape([784, 100]))
b1 = Parameter(float_element_type, Shape([100]))
X3 = Dot(X2, W1) + Broadcast(b1, Shape([bz, 100]), AxisSet({0}))
X4 = relu(X3)
# Affine 2
W2 = Parameter(float_element_type, Shape([100, 10]))
b2 = Parameter(float_element_type, Shape([10]))
X5 = Dot(X4, W2) + Broadcast(b2, Shape([bz, 10]), AxisSet({0}))
# Softmax
Logits = X5
Exp = Exp(Logits)
Max = Reduce(Exp, make_float32_constant(0., [], set()), MaxFn, AxisSet({1}))
MaxBroadcast = Broadcast(Max, Shape([bz, 10]), AxisSet({1}))
Softmax = Exp / MaxBroadcast
# Loss