The official implementation of [CVPR2022] Decoupled Knowledge Distillation https://arxiv.org/abs/2203.08679 and [ICCV2023] DOT: A Distillation-Oriented Trainer https://openaccess.thecvf.com/content/ICCV2023/papers/Zhao_DOT_A_Distillation-Oriented_Trainer_ICCV_2023_paper.pdf
I directly pulled your code and directly used the .yaml settings in your warehouse to run the model. On a single card 4090, it was difficult for me to reproduce the results of the paper. The DKD-Reproduce in the table is the result of my reproduction. It's quite different from the original text. Can you give me some suggestions?
I directly pulled your code and directly used the .yaml settings in your warehouse to run the model. On a single card 4090, it was difficult for me to reproduce the results of the paper. The DKD-Reproduce in the table is the result of my reproduction. It's quite different from the original text. Can you give me some suggestions?