Sergey

This error has been fixed, it works on Windows


@Chilicyy thanks for the answer, in my project I will not use the code from the yolov6 repository, but only use 1 file with the onnx model obtained as a...

> @glenn-jocher hello! > > i was looking in the ultralytics' files of yolov8 and i can't find the loss calculation to modify it. > > How can i manually...

https://github.com/tinyvision/DAMO-YOLO/blob/319572eef367340267ab6ab8ae253527a71d7c3a/damo/detectors/detector.py#L12-L25 you can freeze backbone here

error in you command you ``` !python converter.py -f damoyolo_tinynasL20_T.py -c /content/damo-yolo/damoyolo_tinynasL20_T_436.onnx --batch_size 1 --img_size 640 ``` tutorial `!python tools/converter.py -f configs/damoyolo_tinynasL25_S.py -c damoyolo_tinynasL25_S.pth --batch_size 1 --img_size 640` your need...

make sure that the files are instances_train 2017.json and instances_val2017.json segmentation tags are set to [] (not int) or fix the code here: https://github.com/tinyvision/DAMO-YOLO/blob/319572eef367340267ab6ab8ae253527a71d7c3a/damo/dataset/datasets/coco.py#L92-L103 

try `torchrun tools/train.py -f configs/damoyolo_tinynasL18_Ns.py`

``` diff --git a/tools/train_net.py b/tools/train_net.py index 589ba78..c3eef73 100644 --- a/tools/train_net.py +++ b/tools/train_net.py @@ -149,7 +149,7 @@ if __name__ == "__main__": if args.num_gpus is None: args.num_gpus = torch.cuda.device_count() - extra_sys_path =...

in setup.py replace sklearn with scikit-learn