PRelu - 6 vs 16#
Next section compares an older to a newer version of the same operator after both definition are converted into markdown text.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.
- PRelu6 → PRelu16 +5 -14
PRelu6 → PRelu16
RENAMED
@@ -1 +1 @@
|
|
1
1
|
PRelu takes input data (Tensor<T>) and slope tensor as input, and produces one
|
2
2
|
output data (Tensor<T>) where the function f(x) = slope * x for x < 0,
|
3
3
|
f(x) = x for x >= 0., is applied to the data tensor elementwise.
|
4
|
-
|
5
|
-
**History**
|
6
|
-
- Version 16 adds bfloat16 to the types allowed.
|
7
|
-
This operator supports **unidirectional broadcasting** (tensor slope should be unidirectional broadcastable to input tensor X); for more details please check Broadcasting in ONNX <https://github.com/onnx/onnx/blob/master/docs/Broadcasting.md>_.
|
8
4
|
**Inputs**
|
9
5
|
* **X** (heterogeneous) - **T**:
|
10
6
|
Input tensor
|
11
7
|
* **slope** (heterogeneous) - **T**:
|
12
|
-
Slope tensor.
|
8
|
+
Slope tensor. If Slope is of size 1, the value is sharedacross
|
13
|
-
|
9
|
+
different channels
|
14
10
|
**Outputs**
|
15
11
|
* **Y** (heterogeneous) - **T**:
|
16
|
-
Output tensor
|
12
|
+
Output tensor
|
17
13
|
**Type Constraints**
|
18
14
|
* **T** in (
|
19
|
-
tensor(bfloat16),
|
20
15
|
tensor(double),
|
21
16
|
tensor(float),
|
22
|
-
tensor(float16)
|
17
|
+
tensor(float16)
|
23
|
-
tensor(int32),
|
24
|
-
tensor(int64),
|
25
|
-
tensor(uint32),
|
26
|
-
tensor(uint64)
|
27
18
|
):
|
28
|
-
Constrain input and output types to float
|
19
|
+
Constrain input and output types to float tensors.
|