How to implement a general SSD_mobilenet_v2 model to TensorRT compatible engine

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

Description

I want to know how to implement a general SSD_mobilenet_v2 model to TensorRT compatible engine.I also want to save the engine to be used with Deepstream SDK 4.0

Environment

**TensorRT **: Version5.1.5
GPU Type : NVIDIA T4
Nvidia Driver Version :
CUDA Version release 10.1, V10.1.105 :
nvidia-smi: NVIDIA-SMI 450.36.06 Driver Version: 450.36.06
Operating System + Version : Ubuntu 18.04
Python Version (if applicable) : Python 2.7.17 (default, Apr 15 2020, 17:20:14)
TensorFlow Version (if applicable) :
PyTorch Version (if applicable) :
Baremetal or Container (if container which image + tag) :

2

Hi @GalibaSashi,
You may find the solution in the related post.
Thanks!

1 Like
3

Hi @AakankshaS,
I have a query, based on my understanding for Deepstream SDK, TensorRT 7 is used right. So can you comment whether it is best to use Deepstream SDK 4.0 or 5.0 in terms of reliability, features and support given from NVIDIA. Based on my understanding 5.0 is not as stable as 4.0. Kindly correct me if I am wrong.

4

Hi @GalibaSashi,
You can post your Deepstream related queries on Deepstream SDK Forum.
Thanks!

1 Like