GatherElements - 11 vs 13

Files changed (1) hide show
  1. GatherElements11 → GatherElements13 +5 -8
GatherElements11 → GatherElements13 RENAMED
@@ -1 +1 @@
1
1
  GatherElements takes two inputs data and indices of the same rank r >= 1
2
2
  and an optional attribute axis that identifies an axis of data
3
3
  (by default, the outer-most axis, that is axis 0). It is an indexing operation
4
4
  that produces its output by indexing into the input data tensor at index
5
5
  positions determined by elements of the indices tensor.
6
6
  Its output shape is the same as the shape of indices and consists of one value
7
7
  (gathered from the data) for each element in indices.
8
8
  For instance, in the 3-D case (r = 3), the output produced is determined
9
9
  by the following equations:
10
10
  ::
11
11
  out[i][j][k] = input[index[i][j][k]][j][k] if axis = 0,
12
12
  out[i][j][k] = input[i][index[i][j][k]][k] if axis = 1,
13
13
  out[i][j][k] = input[i][j][index[i][j][k]] if axis = 2,
14
14
  This operator is also the inverse of ScatterElements. It is similar to Torch's gather operation.
15
15
  Example 1:
16
16
  ::
17
17
  data = [
18
18
  [1, 2],
19
19
  [3, 4],
20
20
  ]
21
21
  indices = [
22
22
  [0, 0],
23
23
  [1, 0],
24
24
  ]
25
25
  axis = 1
26
26
  output = [
27
- [
28
- [1, 1],
27
+ [1, 1],
29
- [4, 3],
28
+ [4, 3],
30
- ],
31
29
  ]
32
30
  Example 2:
33
31
  ::
34
32
  data = [
35
33
  [1, 2, 3],
36
34
  [4, 5, 6],
37
35
  [7, 8, 9],
38
36
  ]
39
37
  indices = [
40
38
  [1, 2, 0],
41
39
  [2, 0, 0],
42
40
  ]
43
41
  axis = 0
44
42
  output = [
45
- [
46
- [4, 8, 3],
43
+ [4, 8, 3],
47
- [7, 2, 3],
44
+ [7, 2, 3],
48
- ],
49
45
  ]
50
46
  **Attributes**
51
47
  * **axis**:
52
48
  Which axis to gather on. Negative value means counting dimensions
53
49
  from the back. Accepted range is [-r, r-1] where r = rank(data).
54
50
  **Inputs**
55
51
  * **data** (heterogeneous) - **T**:
56
52
  Tensor of rank r >= 1.
57
53
  * **indices** (heterogeneous) - **Tind**:
58
54
  Tensor of int32/int64 indices, with the same rank r as the input.
59
55
  All index values are expected to be within bounds [-s, s-1] along
60
56
  axis of size s. It is an error if any of the index values are out of
61
57
  bounds.
62
58
  **Outputs**
63
59
  * **output** (heterogeneous) - **T**:
64
60
  Tensor of the same shape as indices.
65
61
  **Type Constraints**
66
62
  * **T** in (
63
+ tensor(bfloat16),
67
64
  tensor(bool),
68
65
  tensor(complex128),
69
66
  tensor(complex64),
70
67
  tensor(double),
71
68
  tensor(float),
72
69
  tensor(float16),
73
70
  tensor(int16),
74
71
  tensor(int32),
75
72
  tensor(int64),
76
73
  tensor(int8),
77
74
  tensor(string),
78
75
  tensor(uint16),
79
76
  tensor(uint32),
80
77
  tensor(uint64),
81
78
  tensor(uint8)
82
79
  ):
83
80
  Constrain input and output types to any tensor type.
84
81
  * **Tind** in (
85
82
  tensor(int32),
86
83
  tensor(int64)
87
84
  ):
88
85
  Constrain indices to integer types