ValueError after training for a few epochs

[Wong-denis]

Hi, I really love your guide and demonstration of instance segmentation using yolov7. I have trained many models successfully. However , everytime I try to train with larger epoch (that's say 1200) it would raise "ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 3 dimension(s) and the array at index 1 has 1 dimension(s)"
It would not raise this error in the first 200 epoch, it usually raise the error at around 300 epoch.
Please tell me if you happen to know how to solve this. I really want to increase my accuracy

[Wong-denis]

solved it! refer to WongKinYiu/yolov7#795 (comment)
According to what they said, the error is happening since mixup has a random probability of being selected as the augmentation to use for a particular step.
It can be solved by using hyp.scratch-low.yaml if mixup is not desirable
Or keep using hyp.scratch-high.yaml but set mixup to 0.0