Darya Vinogradova

I can't provide the code, but I localized the reason of crash - NmsOp from MMDet. However, I don't know how to rewrite it yet. Is that bug fixed?

Basically, is it possible to use nms in tensorrt int8? Could you advice me some working option?

It is an operator from MMDet library. Now I rewrote it in Pytorch and trying to convert to onnx: ``` def nms(boxes: array_like_type, scores: array_like_type, iou_threshold: float, offset: int =...

I prepared a toy example of a mistake in this function. 1. Code of onnx model creation: ``` import torch import numpy as np def nms(boxes, scores, iou_threshold: float, offset:...

I just removed nms layer from my model. I still don't know about good decision :(

onnx file: https://drive.google.com/file/d/1OL7FC4mmGQQqjWAaYLdxcbDfesmSu_S8/view?usp=sharing calibration data: https://drive.google.com/file/d/1ob-tog2o0DDnDlFzN-p35gjLIpbaKKUw/view?usp=sharing pre-built ops: https://drive.google.com/file/d/18GBp1-ounLNqEReDOLchgUKqYzFDnC8F/view?usp=sharing (or you can build this .so as described here: https://github.com/OpenGVLab/InternImage/tree/master/detection)

I suppose, problem is in calibration file. Now I have such a problem `[05/22/2023-06:57:41] [E] Error[1]: [resizingAllocator.cpp::deallocate::105] Error Code 1: Cuda Runtime (an illegal memory access was encountered)` Something is...

Yes, I'm interested in demo