fengjk12138

[Feature] Add full precision context to kfiou loss

2

### What's the feature? ### version dev-1.x ### situation I use kfiou on my dataset. In order to speed up training, I try to use `AmpOptimWrapper`. There is an error...

'torch.dtype' object is not callable

2

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand...

### Bug Description When deploying using docker, an error will be reported if the proxy url uses a domain name. It only works if you change it to a numeric...

在我的统计中，short-term的测试集总共有63000帧左右可用帧，而short-term训练集只有40000帧左右的可用帧，这正常吗，一般训练集的帧数小于测试集的帧数，这合理吗

In the line, https://github.com/zhang-pengyu/HMFT/blob/215cb846fc5da59e4e7810ac92bdc085840b0270/mfDiMP/ltr/models/fusion_model/feature_fusion.py#L31C7-L31C7 , Operations should not be performed on batch_size dim, I think it may be `torch stack` Can you answer my doubts?

readme中只有微调完的权重，请问是否可以分享预训练权重，方便我们在其他数据集上进行微调