anosorae/IRRA

Issues

• Confusion about the IRR module

  #38 opened 8 months ago by shallowzz
  0

• CUHK-PEDES

  #34 opened 9 months ago by LYY123444
  0

• 想问一下想得到主观结果是运行哪个文件呢

  #37 opened 8 months ago by xiaoxiao12580
  0

• The nan error

  #36 opened 8 months ago by tengxiao14
  0

• 请问谁有ICFG-PEDES数据集，救救孩子，发邮件给作者三天了也没有回复

  #35 opened 8 months ago by Gpn-star
  0

• How to specify a particular GPU to train？

  #30 opened 10 months ago by ppsychoo
  1

• Multi gpu training problem

  #23 opened a year ago by Qusijia
  2

• Maybe an error in the original paper?

  #33 opened 9 months ago by ThomasjonLi
  0

• Confusion about cmpm loss and the introduced sdm loss.

  #19 opened a year ago by zhixhan
  1

• Maybe an error

  #32 opened 9 months ago by ThomasjonLi
  0

• a small bug about RSTPReid dataset

  #31 opened 10 months ago by shujielan
  2

• There seems to be a bug at datasets/bases.py#L157.

  #21 opened a year ago by zhixhan
  1

• PROBLEMS REGARDING DIFFRENT SPIT OF DATASET

  #25 opened a year ago by hungphongtrn
  0

• How to Solve multi-GPU Training Problems?

  #22 opened a year ago by Rewinner991130
  1

• There is a mistake that the mlm module that the mask token's output and the whole token

  #29 opened 10 months ago by jssyzsfzy
  1

• 文本全局特征如何获得？

  #28 opened 10 months ago by shams2023
  2

• KeyError: 'mlm_ids'

  #27 opened 10 months ago by yangmengmeng123
  0

• The id loss in the table is not a separate loss, but is trained jointly with the itc loss in the baseline.

  #26 opened a year ago by shams2023
  0

• Is it unfair to use a pre-trained CLIP model compared to some other methods in Table 1?

  #24 opened a year ago by maoSuxi
  0

• A small multi GPU training bug

  #5 opened a year ago by Aotle
  4

• Confusion about the setting of validation dataset

  #20 opened a year ago by YwSama
  1

• About using RN50 as image encoder

  #18 opened a year ago by sunmenmian
  1

• Confusion about the multimodal interaction encoder and mlm task

  #17 opened a year ago by qjyyyy
  3

• Confusion about caption tokens and mlm tokens when using mlm

  #13 opened a year ago by lzfff12
  11

• The accuracy of IDLoss experiment is much lower than 65.33% of the paper

  #16 opened a year ago by xie-xue
  2

• table4. cuhk-pedes rank1 maybe typos~

  #15 opened a year ago by panqihe-zjut
  0

• fix device for ids in _compute_embedding in Evaluator

  #11 opened a year ago by AwePhD
  3

• 关于id_loss在训练过程中不收敛的问题

  #9 opened a year ago by Zplusdragon
  1

• local variable 'num_classes' referenced before assignment

  #12 opened a year ago by CottonCandyZ
  1

• Missing resize_text_pos_embed in clip_model.py

  #10 opened a year ago by AwePhD
  2

• Code for straight forward CLIP fine-tuned to PEDES

  #8 opened a year ago by AwePhD
  2

• cant use best.pth

  #7 opened a year ago by Justhuayu
  2

• about the CUHK-PEDES dataset

  #6 opened a year ago by Justhuayu
  0

• Unexpected keyword argument 'batch_first'

  #4 opened a year ago by LKRCharon
  1

• 请问哪里可以看到论文？

  #3 opened a year ago by LinDixuan
  2

• A small bug

  #2 opened a year ago by YwSama
  1

• About ICFG-PEDES dataset

  #1 opened a year ago by ljwwwiop
  4