Operators
Add
This operator is used to add two tensors.
Cast
Cast Operator casts the given input tensor into the dtype given in the parameter.
Concatenate
Two tensors are concatenetaed on a specific dimension or axis using this operator.
Divide
This operator is used to divide a tensor from another.
Expand Dims
Expand Dims Operator expands the dimensions of the input tensor at the given axis.
Fork
This is basically a copy operator. The input tensor is copied as N output tensors.
Matmul
Performs Matrix Multiplication Operation.
Multiply
This operator is used for multiplication of two tensors.
Reduce
Reduce Operator reduces the input tensor based on the given action one among MEAN, MIN, MAX, SUM, ARGMAX at a given axis.
Reshape
This operator performs the reshape on a given tensor to the new shape specified.
Slice
This operation extracts a slice of size from a tensor input starting from the specified begin location.
Split
This operator splits a tensor into sub tensors on a specified axis. This operation is inverse of Concatenation.
Squeeze
Squeeze Operator removes all the dimensions of size 1 from the shape of a tensor. If axis is provided the operation is performed only on the given axis.
Subtract
This operator is used to subtract a tensor from another.
Tile
Tile Operator creates a new tensor by replicating input multiples times. The output tensor's i'th dimension has input.dims(i) * multiples[i] elements, and the values of input are replicated multiples[i] times along the 'i'th dimension. For example, tiling [a b c d] by [2] produces [a b c d a b c d].
Timing Signal
This Operator adds a bunch of sinusoids of different frequencies.Each channel of the input Tensor is incremented by a sinusoid of a different frequency and phase.
Transpose
Transpose Operation's output dimension i will correspond to the input dimension perm[i]. If perm is not given, it is set to (n-1...0), where n is the rank of the input tensor. Hence by default, this operation performs a regular matrix transpose on 2-D input Tensors. If conjugate is True and a.dtype is either complex64 or complex128 then the values of a are conjugated and transposed.
Unstack
Unpacks the given dimension of a rank-R tensor into rank-(R-1) tensors.