JieShibo/PETL-ViT

Issues

• Issue about training configuration on Binary Adapter with VTAB-1k Dataset

  #19 opened 5 months ago by leoli646
  1

• Few-shot results

  #11 opened a year ago by heekhero
  3

• question about the initialization

  #18 opened a year ago by zheng-chen777
  0

• convpass for multi-gpus training

  #17 opened a year ago by Fanye12
  0

• CIfar100 experimental results query

  #16 opened a year ago by lloo099
  5

• Question about the performance of VTAB-1K with FacT

  #15 opened a year ago by Bazinga699
  4

• The difference between the experimental results on VTAB-1K in your two papers.

  #14 opened a year ago by shachoi
  2

• About the weight initialization for the 'adapter_conv' in Convpass

  #13 opened a year ago by Harry-zzh
  1

• paper link of Revisiting the Parameter Efficiency of Adapters from the Perspective of Precision Redundancy?

  #12 opened a year ago by qgh1223
  1

• Few-Shot Learning dataset

  #10 opened a year ago by RaptorMai
  1

• FacT for Swin

  #9 opened a year ago by RaptorMai
  1

• Full-tuning and Linear accuracy is strange

  #8 opened 2 years ago by sungbinson
  4

• Question about the reported performance.

  #7 opened 2 years ago by luogen1996
  2

• A question for the architecture

  #6 opened 2 years ago by YuchenLiu98
  1

• Could you please share the code for other baseline methods

  #5 opened 2 years ago by Pang-Yatian
  3

• Some questions about SwinTrans edition

  #4 opened 2 years ago by 923714892
  1

• Some questions about the dimension

  #3 opened 2 years ago by zhangbw21
  1

• The number of trainable parameters

  #2 opened 2 years ago by Zhangwenyao1
  1

• When can the code be released?

  #1 opened 2 years ago by RizhaoCai
  1