Train the teacher and student end-to-end in an iterative fashion

[Houssem171]

is there a way to do an iterative training where the teacher can learn from the student's feedback?

[NeelayS]

Hi @Houssem171, we currently do not have such a method implemented in the library. Do you any specific paper/method in mind?

[Houssem171]

the method is mentioned in this paper "Emerging Properties in Self-Supervised Vision Transformers."
link : https://openaccess.thecvf.com/content/ICCV2021/papers/Caron_Emerging_Properties_in_Self-Supervised_Vision_Transformers_ICCV_2021_paper.pdf

[NeelayS]

Unfortunately, we don't have an implementation of this paper in the library and neither is it straightforward for a user to extend the library's methods to quickly implement it.
If you wish, you can create an issue for this and we'll try to get around to implementing it.