TopK - 1 vs 10

Files changed (1) hide show
  1. TopK1 → TopK10 +4 -2
TopK1 → TopK10 RENAMED
@@ -1 +1 @@
1
1
  Retrieve the top-K elements along a specified axis. Given an input tensor of
2
2
  shape [a_1, a_2, ..., a_n, r] and integer argument k, return two outputs:
3
3
  -Value tensor of shape [a_1, a_2, ..., a_{axis-1}, k, a_{axis+1}, ... a_n]
4
4
  which contains the values of the top k elements along the specified axis
5
5
  -Index tensor of shape [a_1, a_2, ..., a_{axis-1}, k, a_{axis+1}, ... a_n] which
6
6
  contains the indices of the top k elements (original indices from the input
7
7
  tensor).
8
+
8
9
  Given two equivalent values, this operator uses the indices along the axis as
9
10
  a tiebreaker. That is, the element with the lower index will appear first.
10
11
  **Attributes**
11
12
  * **axis**:
12
13
  Dimension on which to do the sort.
13
- * **k** (required):
14
- Number of top elements to retrieve
15
14
  **Inputs**
16
15
  * **X** (heterogeneous) - **T**:
17
16
  Tensor of shape [a_1, a_2, ..., a_n, r]
17
+ * **K** (heterogeneous) - **tensor(int64)**:
18
+ A 1-D tensor containing a single positive value corresponding to the
19
+ number of top elements to retrieve
18
20
  **Outputs**
19
21
  * **Values** (heterogeneous) - **T**:
20
22
  Tensor of shape [a_1, a_2, ..., a_{axis-1}, k, a_{axis+1}, ... a_n]
21
23
  containing top K values from the input tensor
22
24
  * **Indices** (heterogeneous) - **I**:
23
25
  Tensor of shape [a_1, a_2, ..., a_{axis-1}, k, a_{axis+1}, ... a_n]
24
26
  containing the corresponding input tensor indices for the top K
25
27
  values.
26
28
  **Type Constraints**
27
29
  * **T** in (
28
30
  tensor(double),
29
31
  tensor(float),
30
32
  tensor(float16)
31
33
  ):
32
34
  Constrain input and output types to float tensors.
33
35
  * **I** in (
34
36
  tensor(int64)
35
37
  ):
36
38
  Constrain index tensor to int64