zkcys001/UDAStrongBaseline

Issues

• 请问是否可以公布weights文件，想基于本工作在模型推理加速方面进行探索

  #33 opened 10 months ago by mookixiao
  0

• 训练到一定epoch停滞的问题

  #15 opened 3 years ago by LeeYushan
  3

• 请问可以使用单卡训练吗？大概需要多少显存？

  #11 opened 3 years ago by LeeYushan
  3

• 训练到一定epoch之后（50+）模型输出、loss会变成nan

  #25 opened 3 years ago by yanty123
  2

• msmt17的训练脚本没有嘛？

  #27 opened 3 years ago by xmy0916
  3

• A question about the case of Market2MSMT

  #32 opened 2 years ago by LeeGeon30
  0

• 跑GLT II的时候出现以下问题

  #31 opened 2 years ago by shylie728
  0

• 跑GLT II 出现以下问题

  #30 opened 2 years ago by shylie728
  0

• 关于几个loss的困惑

  #29 opened 3 years ago by Alex20172023
  0

• A question when I run the code of GLT Stage II

  #28 opened 3 years ago by pumpkin-vera
  0

• 关于resnet50网络架构里面的_build_nonlocal函数

  #24 opened 3 years ago by henrysun1996
  3

• 数据采样的困惑

  #26 opened 3 years ago by muzishen
  0

• 请问baseline是否加了对比学习损失呢？

  #19 opened 3 years ago by yanty123
  1

• 为什么这儿的精度复现出来只有83.3%的top1，

  #23 opened 3 years ago by henrysun1996
  9

• 训练精度不对齐的问题

  #20 opened 3 years ago by Youngboy12
  5

• 关于fine-tune阶段source data 的ID loss 问题

  #18 opened 3 years ago by yanty123
  2

• RuntimeError: cuDNN error: CUDNN_STATUS_EXECUTION_FAILED

  #13 opened 3 years ago by yanchi2021
  1

• GLT_kmeans_duke2market复现很多log

  #21 opened 3 years ago by henrysun1996
  0

• 图片如此

  #22 opened 3 years ago by henrysun1996
  0

• 模型输出后，为什么要删掉一半的特征？

  #17 opened 3 years ago by woocoder
  2

• 关于消融实验

  #16 opened 3 years ago by woocoder
  3

• 数据集问题

  #12 opened 3 years ago by henrysun1996
  1

• RuntimeError: CUDA error: device-side assert triggered

  #10 opened 3 years ago by henrysun1996
  5

• Strong baseline

  #14 opened 3 years ago by clarkhedi
  4

• 三元组损失的形式

  #9 opened 3 years ago by clarkhedi
  2

• missing sbs_traindbscan_unc.py

  #6 opened 3 years ago by yjh576
  1

• The result of GLT much lower than paper's

  #8 opened 3 years ago by woocoder
  2

• GLT_kmeans_duke2market raise StopIteration error

  #7 opened 3 years ago by woocoder
  1

• the code of [Our stronger_baseline + memory bank]

  #4 opened 3 years ago by Wang-pengfei
  2

• source domain pre-training

  #5 opened 3 years ago by hahapt
  5

• MMT results

  #3 opened 4 years ago by SikaStar
  1

• Very bad result for cross-domain reid

  #2 opened 4 years ago by shuxjweb
  2

• No resnet.py and resnet_ibg.py in your UDAsbs.models directory.

  #1 opened 4 years ago by justopit
  0