youngjae-you

@ryouchinsa Well, in the end, it depends on how the train code parses it, but I'm curious about whether that method is efficient.

@AyushExel I'm talking about the comparison of mAP numerical information for small area output from **cocoeval summarize.**

1.37.0 still occrs error

@imzhangyd What do you mean by `train`? Isn't Bytetrack simply a data association method?

그리고 추가적으로 BUILD_TYPE=DEBUG, RELEASE차이는 라이브러리 사이즈 차이라고 보면 되나요? 라이브러리 속도 사이즈와는 무관한거죠?

@marcoslucianops https://github.com/marcoslucianops/DeepStream-Yolo/tree/b2c4bee8dcbc4729f1bd13b598a574228ff10c27 Can you tell me what changes were applied in this commit, the master commit, and the yolov5 side?

@marcoslucianops I am moving from the commit mentioned in the text to the current master commit. However, to build the trt engine in the onnx model, in the past, the...

@marcoslucianops It seems that the default build setting is **_explicit batch._**

It's related issue by mentioned me (#1029

@githubzhao735 As of today, 0.30.alpha seems to be no different from master