Unable to inference a trt model in jetson nano/ xavier nx

Autonomous Machines Jetson & Embedded Systems Jetson TX2
,
1

Description

I am new to Deep stream. I had created a custom model for dog and cat detection of detectnet_v2. I had converted etlt file to trt.int8 file . I had tried to find some resource in google but unable to find any helpful resource.

resnet18_cat_dog.etlt
resnet18_car_dog.trt.int8

I want to do inference in jetson devices. can anyone guide me to some resources which i can use to do inference.

Environment

Jetson Nano
Jerson Xavier NX

2

Hi,
This looks like a Jetson issue. Please refer to the below samlples in case useful.

For any further assistance, we recommend you to raise it to the respective platform from the below link

Thanks!

5

Hi,

Please noted that Nano doesn’t support INT8 operation due to hardware limitations.
You will need a Xavier/XavierNX for this feature or use fp32/fp16 mode instead.

More, since TensorRT is not portable, you will need to convert the TensorRT engine on the Jetson directly.
You can find a step-by-step document below:
https://docs.nvidia.com/tao/tao-toolkit/text/deepstream_tao_integration.html

Thanks.

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.