s2anet支持使用其他分类损失吗？我尝试了eql_loss(用于处理长尾现象），但是训练时，报错如下：如果使用别的损失，需要注意哪些地方吗，希望您可以给出一些建议。

[xc-chengdu]

[csuhan]

Maybe you can add 'EQLv2' to this line so that we do not use another sampler.

s2anet/mmdet/models/anchor_heads_rotated/s2anet_head.py

Line 64 in d287096

self.sampling = loss_odm_cls['type'] not in ['FocalLoss', 'GHMC']

[xc-chengdu]

Thanks for your suggestion, I will try it.

[xc-chengdu]

Hello, I would like to ask again, have you tried using the non-cuda version of focal loss? Why is there such a big difference between the two trained models?