IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update FakeQuantize spec (#6608)

Browse Source 
* update FakeQuantize spec

* remove `**Default value**: None` for required attribute
This commit is contained in:
Patryk Elszkowski 2021-07-15 07:45:37 +02:00 committed by GitHub
parent c14edd4134
commit ca9e495e13
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
docs/ops/quantization/FakeQuantize_1.md
View File

@ -36,29 +36,34 @@ else:
* **Description**: specifies rules used for auto-broadcasting of input tensors. * **Description**: specifies rules used for auto-broadcasting of input tensors.
* **Range of values**: * **Range of values**:
* *none* - no auto-broadcasting is allowed, all input shapes should match * *none* - no auto-broadcasting is allowed, all input shapes should match
* *numpy* - numpy broadcasting rules, aligned with ONNX Broadcasting. Description is available in <a href="https://github.com/onnx/onnx/blob/master/docs/Broadcasting.md">ONNX docs</a> * *numpy* - numpy broadcasting rules, description is available in [Broadcast Rules For Elementwise Operations](../broadcast_rules.md),
* *pdpd* - PaddlePaddle-style implicit broadcasting. * *pdpd* - PaddlePaddle-style implicit broadcasting, description is available in [Broadcast Rules For Elementwise Operations](../broadcast_rules.md).
* **Type**: string * **Type**: string
* **Default value**: "numpy" * **Default value**: "numpy"
* **Required**: *no* * **Required**: *no*
**Inputs**: **Inputs**:
* **1**: `X` - multidimensional input tensor of floating type to be quantized. Required. * **1**: `X` - tensor of type *T_F* and arbitrary shape. **Required.**
* **2**: `input_low` - minimum limit for input value. The shape must be broadcastable to the shape of *X*. Required. * **2**: `input_low` - tensor of type *T_F* with minimum limit for input value. The shape must be broadcastable to the shape of *X*. **Required.**
* **3**: `input_high` - maximum limit for input value. Can be the same as `input_low` for binarization. The shape must be broadcastable to the shape of *X*. Required. * **3**: `input_high` - tensor of type *T_F* with maximum limit for input value. Can be the same as `input_low` for binarization. The shape must be broadcastable to the shape of *X*. **Required.**
* **4**: `output_low` - minimum quantized value. The shape must be broadcastable to the shape of *X*. Required. * **4**: `output_low` - tensor of type *T_F* with minimum quantized value. The shape must be broadcastable to the shape of *X*. **Required.**
* **5**: `output_high` - maximum quantized value. The shape must be broadcastable to the of *X*. Required. * **5**: `output_high` - tensor of type *T_F* with maximum quantized value. The shape must be broadcastable to the of *X*. **Required.**
**Outputs**: **Outputs**:
* **1**: `Y` - resulting tensor with shape and type matching the 1st input tensor *X*. * **1**: output tensor of type *T_F* with shape and type matching the 1st input tensor *X*.
**Types**
* *T_F*: any supported floating point type.
**Example** **Example**
```xml ```xml
<layer … type="FakeQuantize"…> <layer … type="FakeQuantize"…>
<data levels="2"/> <data levels="2"/>