BatchToSpace for 4-D tensors of type T.

BatchToSpace for N-D tensors of type T.

Bitcasts a tensor from one type to another without copying data.

Return the shape of s0 op s1 with broadcast.

Checks a tensor for NaN and Inf values.

Concatenates tensors along one dimension.

Identity op for gradient debugging.

DepthToSpace for tensors of type T.

Dequantize the 'input' tensor into a float Tensor.

Returns a diagonal tensor with a given diagonal values.

Returns the diagonal part of the tensor.

Computes the (possibly normalized) Levenshtein Edit Distance.

Inserts a dimension of 1 into a tensor's shape.

Extract patches from images and put them in the "depth" output dimension.

Fake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same type.

Compute gradients for a FakeQuantWithMinMaxArgs operation.

Fake-quantize the 'inputs' tensor of type float via global float scalars min

Compute gradients for a FakeQuantWithMinMaxVars operation.

Fake-quantize the 'inputs' tensor of type float and one of the shapes: [d],.

Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.

Creates a tensor filled with a scalar value.

Gather slices from params according to indices.

Gather slices from params into a Tensor with shape specified by indices.

Gather slices from params axis axis according to indices.

Return a tensor with the same shape and contents as the input tensor or value.

Returns a list of tensors with the same shapes and contents as the input.

Returns immutable tensor from memory region.

Computes the inverse permutation of a tensor.

Copy a tensor setting everything outside a central band in each innermost matrix.

Returns a batched diagonal tensor with a given batched diagonal values.

Returns the batched diagonal part of a batched tensor.

Returns a batched matrix tensor with new batched diagonal values.

Pads a tensor with mirrored values.

Returns a one-hot tensor.

Returns a tensor of ones with the same shape and type as x.

Pads a tensor with zeros.

Pads a tensor.

Concatenates a list of N tensors along the first dimension.

A placeholder op for a value that will be fed into the computation.

A placeholder op that passes through input when its output is not fed.

An identity op that triggers an error if a gradient is requested.

Quantizes then dequantizes a tensor.

Quantizes then dequantizes a tensor.

Quantize the 'input' tensor of type float to 'output' tensor of type 'T'.

Concatenates quantized tensors along one dimension.

Quantized Instance normalization.

Reshapes a quantized tensor as per the Reshape op.

Returns the rank of a tensor.

Reshapes a tensor.

Assignvalue to the sliced l-value reference of ref.

Reverses specific dimensions of a tensor.

Reverses variable length slices.

Scatter updates into a new (initially zero) tensor according to indices.

Applies sparse addition to input using individual values or slices.

Computes the difference between two lists of numbers or strings.

Returns the shape of a tensor.

Returns shape of tensors.

Returns the size of a tensor.

Return a slice from 'input'.

SpaceToBatch for 4-D tensors of type T.

SpaceToBatch for N-D tensors of type T.

SpaceToDepth for tensors of type T.

Splits a tensor into num_split tensors along one dimension.

Splits a tensor into num_split tensors along one dimension.

Removes dimensions of size 1 from the shape of a tensor.

Packs a list of N rank-R tensors into one rank-(R+1) tensor.

Stops gradient computation.

Return a strided slice from input.

Assignvalue to the sliced l-value reference of ref.

Returns the gradient of StridedSlice.

Constructs a tensor by tiling a given tensor.

Shuffle dimensions of x according to a permutation.

Finds unique elements in a 1-D tensor.

Finds unique elements in a 1-D tensor.

Unpacks a given dimension of a rank-R tensor into num rank-(R-1) tensors.

Returns locations of true values in a boolean tensor.

Returns a tensor of zeros with the same shape and type as x.