Alexander Mamaev

@Poulinakis-Konstantinos I think may be it's happens with recent version of pytorch/yolov5. Can you check your model with yolov5 version at Aug 2021?

@adrianosantospb @Poulinakis-Konstantinos I am not a creator of the convertor, I use `trtexec` util from this repo https://github.com/NVIDIA/TensorRT I think, that may few reasons of problem. 1. The first is...

Probably something is wrong with the model deserialization, that's happening [here](https://github.com/alxmamaev/jetson_yolov5_tensorrt/blob/a39aeac41dfeea055297169a8a90082c44bf9ea7/yolov5_trt/trt_wrapper.py#L43). You may check TensorRT docs for this method, I think that is a problem with the model conversion.

@little-siskin Hello, you are do not provide an example of the model wrong output. By the way, it may depend on your model or environment, the inference code is simple...