T
- data type for values
output@Operator(group="nn") public final class TopK<T extends TNumber> extends RawOp
k
largest elements for the last dimension.
If the input is a vector (rank-1), finds the k
largest entries in the vector
and outputs their values and indices as vectors. Thus values[j]
is the
j
-th largest entry in input
, and its index is indices[j]
.
For matrices (resp. higher rank input), computes the top k
entries in each
row (resp. vector along the last dimension). Thus,
values.shape = indices.shape = input.shape[:-1] + [k]
If two elements are equal, the lower-index element appears first.
Modifier and Type | Class and Description |
---|---|
static class |
TopK.Inputs<T extends TNumber> |
static class |
TopK.Options
Optional attributes for
TopK |
Modifier and Type | Field and Description |
---|---|
static String |
OP_NAME
The name of this op, as known by TensorFlow core engine
|
Modifier and Type | Method and Description |
---|---|
static <T extends TNumber> |
create(Scope scope,
Operand<T> input,
Operand<TInt32> k,
TopK.Options... options)
Factory method to create a class wrapping a new TopKV2 operation.
|
Output<TInt32> |
indices()
Gets indices.
|
static TopK.Options |
sorted(Boolean sorted)
Sets the sorted option.
|
Output<T> |
values()
Gets values.
|
public static final String OP_NAME
public TopK(Operation operation)
@Endpoint(describeByClass=true) public static <T extends TNumber> TopK<T> create(Scope scope, Operand<T> input, Operand<TInt32> k, TopK.Options... options)
T
- data type for TopKV2
output and operandsscope
- current scopeinput
- 1-D or higher with last dimension at least k
.k
- 0-D. Number of top elements to look for along the last dimension (along each
row for matrices).options
- carries optional attribute valuespublic static TopK.Options sorted(Boolean sorted)
sorted
- If true the resulting k
elements will be sorted by the values in
descending order.public Output<T> values()
k
largest elements along each last dimensional slice.Copyright © 2015–2022. All rights reserved.