YU1ut/MixMatch-pytorch

Issues

• At what point does the program stop running

  #47 opened 3 years ago by Enderfga
  2

• train_loss

  #40 opened 3 years ago by Wanghd-MVP
  1

• Question about the Meaning of Accuracy in the Table Results

  #14 opened 2 years ago by EveDong
  4

• I use mixmatch to train my own data set, which is simpler than cifar10 image. After running several epoch, the loss function of training and trainstats will become larger, and the accuracy of test set will be reduced

  #36 opened 4 years ago
  3

• why use .detach() here

  #44 opened 3 years ago by hzhz2020
  0

• Question about forward pass for unlabel samples in training process.

  #43 opened 3 years ago by DingYuan0118
  3

• Unlabelled batch size

  #39 opened 4 years ago by AlexTS1980
  5

• interleave function

  #41 opened 3 years ago by 07Agarg
  1

• Performance declining by replacing data augmentation with pyotrch function.

  #19 opened 5 years ago by yichuan9527
  3

• How to apply it to my own dataset？

  #34 opened 4 years ago
  3

• How to adjust α, λ u in our own data set? Any tips

  #37 opened 4 years ago
  2

• I can not find the class 'Bar' in the code. Anyone could help me please!

  #30 opened 4 years ago by Alagirl
  2

• wrong transform order

  #35 opened 4 years ago by yuanzhanghu
  1

• A question on lambda_u

  #33 opened 4 years ago by zhaozhengChen
  2

• The description of the parameter is not clear

  #32 opened 4 years ago by ustcjerry
  1

• Qustion about the update method of ema_model

  #16 opened 5 years ago by sanyouwu
  5

• Have you tested the codes with other models?

  #7 opened 6 years ago by shuxjweb
  8

• bug

  #27 opened 5 years ago by afterall204168
  2

• Excellen work. I have run the code, and best acc is 93.78.

  #28 opened 5 years ago by chuxiang93
  0

• What does the interleave method do

  #20 opened 5 years ago by atinghosh
  2

• why unlabel numbers two times than label?

  #26 opened 5 years ago by daixiangzi
  2

• Parameter assignment bug.

  #24 opened 5 years ago by dddzg
  1

• No weight decay?

  #22 opened 5 years ago by vijosam
  2

• Memory leak

  #21 opened 5 years ago by HHHedo
  0

• Anybody try this method on NLP tasks?

  #17 opened 5 years ago by xvshiting
  0

• The bias term of convolution in WideResNet

  #10 opened 6 years ago by blankWorld
  1

• Importance to update the lambda slowly

  #9 opened 5 years ago by wang3702
  4

• Implementation of the Lamba_u is not correct

  #6 opened 5 years ago by wang3702
  3

• The question about batch_size/val_iteration/lr.

  #11 opened 5 years ago by xiaopingzeng
  2

• It seems the mean teacher does not contribute to the loss.

  #13 opened 5 years ago by jonkoi
  2

• Question about the efficiency of parameter T

  #15 opened 5 years ago by sanyouwu
  1

• Please add license

  #12 opened 5 years ago by matthew-jurewicz
  0

• about interleave labeled and unlabed samples

  #5 opened 6 years ago by blankWorld
  6

• About the settings of parameters and CIFAR100

  #3 opened 6 years ago by Fire-friend
  4

• This implementation is NOT MixMatch but Mean Teacher

  #4 opened 6 years ago by fperdigon
  1

• Is my data wrong?

  #2 opened 6 years ago by sailor88128
  3

• did you get similar results to the paper?

  #1 opened 6 years ago by AlphaQi
  2