 |
TensorRT
7.2.1.6
NVIDIA TensorRT
|
|
|
TensorRT
7.2.1.6
NVIDIA TensorRT
|
|
Table Of Contents
This plugin applies the Gelu activation x * Phi(x), where Phi is the Gaussian cdf, approximated by: 0.5 * (1 + tanh(sqrt(2 / M_PI) * (x + 0.044715 * x^3))). Optionally adds a bias vector before the activation.
The geluPlugin takes one input; input.
input input is a tensor with shape [S, B, E] where B is the batch size.
The geluPlugin generates the following output:
output output is a tensor with shape [S, B, E] where B is the batch size.
geluPlugin has plugin creator class GeluPluginDynamicCreator and plugin class CustomGeluPluginDynamic.
The parameters are defined below and consists of the following attributes:
| Type | Parameter | Description |
|---|---|---|
int | type_id | Integer encoding the DataType (0: FP32, 1: FP16) |
Weights | bias | Optional bias parameter. Shape [1, 1, E] |
For terms and conditions for use, reproduction, and distribution, see the TensorRT Software License Agreement documentation.
November 2019 This is the first release of this README.md file.
This plugin only supports GPUs with compute capability >= 7.0. For more information see the CUDA GPU Compute Capability Support Matrix
1.8.17