Alexey

Do you do it locally or on Google Colab? Can you successfully convert and run onnx model for yolov7_training.pt model ? https://github.com/WongKinYiu/yolov7/releases/download/v0.1/yolov7_training.pt Could you provide your model?

Yes, this seems to be the reason. Which Conf-thresh value does work well for you?

Try to use this fix for Validation: https://github.com/WongKinYiu/yolov7/pull/368

YOLOv7 inference time: * 6ms on GPU V100 (134 TFlops-TC) * 12ms on GPU T4 (65 TFlops-TC) It is very strange that you achieve 14ms on GPU A100 (310 TFlops-TC),...

@ersheng-ai * Did you use **converted** or **trained** weights on Pytorch? * Do you use keeping aspect ration of image during resizing? YOLOv4 doesn't keep aspect ration by default.

@ersheng-ai Try to use resizing without keeping-aspect-ratio/padding-zeros, will accuracy be better?

@ersheng-ai * github.com/AlexeyAB/darknet - 47.1% AP - 71.0% AP50 * Tianxiaomo/pytorch-YOLOv4 - 46.6% AP (**-0.5**) - 70.3% AP50 (**-0.7**) So there is no big drop in accuracy. Great! Do you...

@maudzung > Hence, I haven't taken advantage of CIoU or GIoU loss for optimization. Did you try CIoU/GIoU for training with 3D-bboxes and it didn't increase accuracy? > I'm trying...

@maudzung Hi, I use: ``` lgiou_scale = 0.07 lobj_scale = 1.0 lcls_scale = 1.0 ``` ---- Also you can try ``` lgiou_scale = 0.05 lobj_scale = 1.0 lcls_scale = 0.6...

What do you mean? I use: ```cpp if (truth) { // for object delta_bbox[i] = giou_delta[i] * lgiou_scale; delta_objectness = (1 - output[obj_index]) * lobj_scale; for(int k = 0; k...