LiJiahang

i have not solved it yet，and im confused too. I think yolo format datasets use .txt as their labels and all the things you have to do is to organzie...

ok i'll try it later .Thanks a lot man. ---Original--- From: "Mahathi ***@***.***> Date: Tue, Jul 5, 2022 13:27 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [meituan/YOLOv6] AssertionError: ../coco/images/train2017 is...

you may have to write a python script to generate .txt file from .xml annotations. ---Original--- From: "Johnny ***@***.***> Date: Thu, Jul 7, 2022 17:00 PM To: ***@***.***>; Cc: ***@***.******@***.***>;...

感谢您的回答，现在我明白了原因。另外有一点疑问是我在复现您的Cityscapes数据集上的结果时，按照您的预训练权重进行训练，但是无法载入您的[Mapillary]权重，在载入时会报错：Attention中的relative position bias的size错误，但是使用您的另一个beitv2-l的权重的时候却可以载入，因此想请问您是不是在cityscapes数据集上的实验时对网络参数进行了修改，比如对rel-pos-bias的使用等   s/103312600/820b9b38-900d-4ee5-a249-e25f2e242788)