SwinTransformer/Feature-Distillation

Issues

• Was Whitening implemented incorrectly?

  #14 opened 2 years ago by jsrdcht
  1

• Finetuned checkpoint of FD-SwinV2-G

  #16 opened a year ago by tung-nd
  0

• What is the difference between distilled model and fine-tuned model?

  #15 opened a year ago by shenyehui
  0

• resnet as student network and CLIP as teacher network

  #13 opened 2 years ago by gqingc
  0

• Can you share the logs for the released models?

  #12 opened 2 years ago by yxchng
  0

• Distilling features of SwinV2-L to SwinV2-T

  #10 opened 2 years ago by yopknopixx
  1

• Which part is teacher in class FD?

  #11 opened 2 years ago by xs1997zju
  0

• About Loss

  #7 opened 2 years ago by Kathrine94
  0

• 哈喽

  #9 opened 2 years ago by ross-Hr
  0

• model release

  #8 opened 2 years ago by futureisatyourhand
  0

• Didn't get the purpose of the method

  #3 opened 2 years ago by NikAleksFed
  3

• can distilling recursively improve the performance?

  #6 opened 2 years ago by ray-lee-94
  2

• When will you release the code of Contrastive Learning Rivals Masked Image Modeling in Fine-tuning via Feature Distillation?

  #4 opened 2 years ago by geekinglcq
  2

• Where do the properties come from?

  #5 opened 2 years ago by jsrdcht
  0

• Are the Teacher and Student the same pretrained model with the same initialization?

  #2 opened 2 years ago by zhangxinyu-xyz
  2

• Code Release

  #1 opened 2 years ago by abhigoku10
  0