Ming Kang

Results 85 comments of Ming Kang

There is a eror: "cannot import models.VovShuffleRep"

Sorry for bugs. Please download the updated source code.

There is a eror: "cannot import models.VovShuffleRep"

> Dear blogger, hello, I accidentally came into contact with your published paper and your algorithm, I am very interested in it, but when I pulled its code to the...

There is a eror: "cannot import models.VovShuffleRep"

> VovShuffleRep Please redownload code and run it. There isn't this problem on new version.

There is a eror: "cannot import models.VovShuffleRep"

Solved. This code fits to a suitable running environment. The requirements and recommended running environment were added. We've known that it doesn't work on Python 3.9 + Torch 2.0 or...

RuntimeError: max_pool3d_with_indices_backward_cuda does not have a deterministic implementation, but you set 'torch.use_deterministic_algorithms(True)'.

Did you configure your GPU Device 0?

Trained Weights

[yolov5l-seg.pt](https://github.com/mkang315/ASF-YOLO/blob/main/yolov5l-seg.pt) that contains pretrained weights has already been uploaded.

Trained Weights

Sorry, I misunderstood what you mean. The file yolov5l-seg.pt contain the initial weights that I described in the paper and isn't the pretrained weights of ASF-YOLO. I know you want...

a problem of figure

Thanks for your reminder. I'll update it soon.

RuntimeError: max_pool3d_with_indices_backward_cuda does not have a deterministic implementation, but you set 'torch.use_deterministic_algorithms(True)'.

I suppose your problem is related to hyperparameters that are set in [./data/hyps/hyp.scratch-low.yaml](https://github.com/mkang315/ASF-YOLO/blob/main/data/hyps/hyp.scratch-low.yaml) because the error message can be tracked in [./train.py](https://github.com/mkang315/ASF-YOLO/blob/main/train.py) line 69: def train(hyp, opt, device, callbacks): #...

RuntimeError: max_pool3d_with_indices_backward_cuda does not have a deterministic implementation, but you set 'torch.use_deterministic_algorithms(True)'.

Please look at [YOLOv5 Multi-GPU Tutorial](https://github.com/ultralytics/yolov5/issues/475) and follow it.