Knowledge distillation: A good teacher is patient and consistent

tensorflow:
https://github.com/google-research/big_vision/tree/main/big_vision/configs/proj/distill

Do you have plans to open source the distillation code?