Table Of Contents
The Clip plugin restricts the values of a tensor to the range [clipMin, clipMax]. Values below this range are set to clipMin, values above this range are set to clipMax, and values within this range are unchanged.
The Clip plugin can be used to implement the ReLU6 activation function, which clips the input tensor values to the range [0, 6].
The Clip plugin takes one input. The input can have any shape.
The Clip plugin generates one output. The output will have the same shape as the input.
The Clip plugin supports the data types float32, float16, and int32. The data types int8, uint8, and bool are not supported.
This plugin works for network with graph node named Clip_TRT. This is also the plugin name that should be used when getting the ClipPluginCreator from the Plugin Registry.
The Clip plugin has plugin creator class ClipPluginCreator and plugin class ClipPlugin.
The parameters are defined below and consists of the following attributes:
| Type | Parameter | Description |
|---|---|---|
float32 |
clipMin |
The lower bound of the range. |
float32 |
clipMax |
The upper bound of the range. |
For terms and conditions for use, reproduction, and distribution, see the TensorRT Software License Agreement documentation.
September 2022
This is the first release of this README.md file.
There are no known issues in this plugin.