Tile - 1 vs 13

Next section compares an older to a newer version of the same operator after both definition are converted into markdown text. Green means an addition to the newer version, red means a deletion. Anything else is unchanged.

Files changed (1) hide show
  1. Tile1 → Tile13 +9 -10
Tile1 → Tile13 RENAMED
@@ -1 +1 @@
1
- Repeat the elements of a tensor along an axis.
1
+ Constructs a tensor by tiling a given tensor.
2
+ This is the same as function tile in Numpy, but no broadcast.
3
+ For example A = [[1, 2], [3, 4]], B = [1, 2], tile(A, B) = [[1, 2, 1, 2], [3, 4, 3, 4]]
2
4
  ### Inputs
3
5
  - **input** (heterogeneous) - **T**:
4
6
  Input tensor of any shape.
5
- - **tiles** (heterogeneous) - **T**:
7
+ - **repeats** (heterogeneous) - **T1**:
8
+ 1D int64 tensor of the same length as input's dimension number, includes numbers of repeated copies along input's dimensions.
6
- Number of repeated copies to make of the input tensor.
7
- - **axis** (heterogeneous) - **T**:
8
-
9
- Axis along which to repeat.
10
9
  ### Outputs
11
10
  - **output** (heterogeneous) - **T**:
12
- Output tensor of same shape and type as input.
11
+ Output tensor of the same dimensions and type as tensor input. output_dim[i] = input_dim[i] * repeats[i]
13
12
  ### Type Constraints
14
- * **T** in ( tensor(double), tensor(float), tensor(float16) ):
13
+ * **T** in ( tensor(bfloat16), tensor(bool), tensor(complex128), tensor(complex64), tensor(double), tensor(float), tensor(float16), tensor(int16), tensor(int32), tensor(int64), tensor(int8), tensor(string), tensor(uint16), tensor(uint32), tensor(uint64), tensor(uint8) ):
15
- Constrain input types to float tensors.
14
+ Constrain input and output types to all tensor types.
16
15
  * **T1** in ( tensor(int64) ):
17
- Constrain tiles and axis's type to int64 tensors.? -------------
16
+ Constrain repeat's type to int64 tensors.? +++++