assert len(indices) == len(self)

[xcx121]

Thank you very much for your work! I encountered the issue shown in the picture: assert len(indices) == len(self) while trying to reproduce your work using my own dataset. I have tried various methods to solve this problem, such as modifying the learning rate; changing samples_per_gpu and workers_per_gpu; and adjusting sample_ratio. However, none of these solutions worked.

I was able to run the configs/consistent-teacher/consistent_teacher_r50_fpn_coco_180k_10p.py code successfully using the COCO dataset with your provided config. But when I switch to my own dataset, the aforementioned issue arises. I have modified num_classes and the categories accordingly.

Could you please advise me on how to resolve this issue? I hope you can provide me with a solution. Thank you!

【非常感谢你们的工作！

我在用自己的数据集复现你们的工作时出现了图片所示的问题： assert len(indices) == len(self)，我尝试了用各种方法去解决这个问题，比如，修改学习率；修改samples_per_gpu和workers_per_gpu；修改sample_ratio。可是都没有解决这个问题。

我用coco数据集可以按照宁提供的config，跑通configs/consistent-teacher/consistent_teacher_r50_fpn_coco_180k_10p.py代码。可是改到我自己的数据集就会出现上述问题，我是修改了num_classessses和类别。

请问我该如何解决这个问题呢，希望您能够给我一个解决方案，谢谢！】