-
How to add more layers to the YOLOv6 architecture? I have a bad quality image dataset & I want to add a few layers to handle image quality. Any suggestions on where and how to add layers to the networ…
-
I think there is a bug in residual connections in inference.
In inferer.py, line 56:
self.model_switch(self.model, self.img_size)
model_switch is supposed to remove residual convolutions and fold…
-
Hi,
Great work!
I've two questions,
1. How to perform inference for instance segmentation using pre-trained model and get the masks?
2. I would like to train instance segmentation model, can you…
-
yolov6.pt中似乎保存了整个模型的结构，一旦在load的时候找不到对应的文件，将引发报错，我认为这是一种很不好用的写法，也可能因为我太无知了吧，总之，很奇怪，我认为的权重就是应该超脱于模型而存在，而不是存在上述的这种耦合。感谢您做出的贡献，但是是不是可能存在更好的写法去保存模型以支持rep-model呢？
如有叨扰，实属得罪
-
Dear Meituan Team,
I am encoutering very low / almost zero evaluation mAP results when running COCO128 tiny datase, whereas using the same setttings of YOLOv5 can already give more than 60% mAP.
…
-
could the author provide the config of large size model?
-
-
I am training on custom dataset
the yaml file is written in this way
![Screenshot from 2022-06-28 14-49-54](https://user-images.githubusercontent.com/18230011/176161465-c631aae4-016c-40da-a19d-74d…
-
how to view mAP results. all classes and every class
-
`Training start...
Epoch iou_loss l1_loss obj_loss cls_loss
0%| | 0/82 [00:00