Dropout - 6 vs 7¶
Next section compares an older to a newer version of the same operator after both definition are converted into markdown text.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.
- Dropout6 → Dropout7 +3 -6
Dropout6 → Dropout7
RENAMED
@@ -1 +1 @@
|
|
1
1
|
Dropout takes one input data (Tensor<float>) and produces two Tensor outputs,
|
2
2
|
output (Tensor<float>) and mask (Tensor<bool>). Depending on whether it is in
|
3
3
|
test mode or not, the output Y will either be a random dropout, or a simple
|
4
4
|
copy of the input. Note that our implementation of Dropout does scaling in
|
5
5
|
the training phase, so during testing nothing needs to be done.
|
6
|
+
This operator has **optional** inputs/outputs. See [ONNX IR](https://github.com/onnx/onnx/blob/main/docs/IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument's name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.
|
6
7
|
### Attributes
|
7
|
-
* **is_test - INT** (default is '0'):
|
8
|
-
|
9
|
-
(int, default 0) if nonzero, run dropout in test mode where the output is simply Y = X.
|
10
|
-
|
11
8
|
* **ratio - FLOAT** (default is '0.5'):
|
12
|
-
|
9
|
+
The ratio of random dropout
|
13
10
|
### Inputs
|
14
11
|
- **data** (heterogeneous) - **T**:
|
15
12
|
The input data as Tensor.
|
16
13
|
### Outputs
|
17
14
|
Between 1 and 2 outputs.
|
18
15
|
- **output** (heterogeneous) - **T**:
|
19
16
|
The output.
|
20
17
|
- **mask** (optional, heterogeneous) - **T**:
|
21
|
-
The output mask.
|
18
|
+
The output mask.
|
22
19
|
### Type Constraints
|
23
20
|
* **T** in ( tensor(double), tensor(float), tensor(float16) ):
|
24
21
|
Constrain input and output types to float tensors.
|