Yolox-nano with deepstream on nvidia jetson - Githubissues

iwatake2222 / play_with_tensorrt

Sample projects for TensorRT in C++

Apache License 2.0

186 stars 33 forks source link

Yolox-nano with deepstream on nvidia jetson #11

Closed tulbureandreit closed 2 years ago

tulbureandreit commented 2 years ago

NVIDIA Jetson AGX Xavier [16GB]
- Jetpack 4.6 [L4T 32.6.1]
- NV Power Mode: MODE_30W_ALL - Type: 3
- jetson_stats.service: active
Libraries:
- CUDA: 10.2.300
- cuDNN: 8.2.1.32
- TensorRT: 8.0.1.6
- Visionworks: 1.6.0.501
- OpenCV: 4.5.1 compiled CUDA: YES
- VPI: ii libnvvpi1 1.1.15 arm64 NVIDIA Vision Programming Interface library
- Vulkan: 1.2.70

Project Name

pj_xxx

Issue Details

I cannot run yolox-nano or yolox-tiny with deepstream on my Jetson. Everytime I try to run them I get a segmentation fault or my terminal freezes. But I can run yolox-s, at 1 FPS, slowly, but I does run.

I convert them from torch to TRT the same way and when doing inference with just TRT, all 3 models work.

How to Reproduce

Convert original yolox-nano, with torch2trt and then Try to use it with deepstream 6.0

iwatake2222 commented 2 years ago

This library is not for DeepStream. I have no idea about DeepStream.