Trtexc specifies the precision for the specified layer

AI & Data Science Deep Learning (Training & Inference) TensorRT
,
1

Description

When I use tensorrt8.6 to convert the onnx model, trtexc can use the layerPrecisions command to constrain the precision of the specified layer, but tensrrt8.2.1 does not have this command. How can I constrain the precision of the specified layer using a lower version of tensorrt8?

Environment

TensorRT Version: tensorrt8.6.1 / tensorrt8.2.1
GPU Type: rtx3060
Nvidia Driver Version: 546.80
CUDA Version: 11.4 /12.3
CUDNN Version:
Operating System + Version:
Python Version (if applicable):
TensorFlow Version (if applicable):
PyTorch Version (if applicable): 12.1
Baremetal or Container (if container which image + tag):

5

Hi @250854911 ,
We always encourgae to use the latest of the release,
However on the pointer, i am checking internally if thats feasible.

Thanks