khanrc/pt.darts

Issues

• very slow in multi-gpu

  #29 opened 5 years ago by YoungDav
  5

• Why is it displayed “NaN or Inf found in input tensor.”

  #47 opened a year ago by iGuojie
  0

• how to use the visualize.py

  #44 opened 3 years ago by try-gx
  1

• what does the "auxiliary_head" mean?

  #45 opened 3 years ago by Sameul-kaji
  0

• batch_size issue

  #40 opened 4 years ago by buttercutter
  2

• Hessian computation issue (seems like there is a bug)

  #43 opened 3 years ago by MartinXPN
  0

• How to train own dataset!

  #42 opened 3 years ago by Kirito-520
  1

• AttributeError: 'CIFAR10' object has no attribute 'train_data'

  #11 opened 6 years ago by leidaguo
  5

• Questions about DARTS

  #41 opened 4 years ago by buttercutter
  0

• low test accuracy in FashionMNIST

  #36 opened 5 years ago by Liang-yc
  0

• Run augment.py and out of memory after one epoch

  #39 opened 4 years ago by weihui98
  0

• Search Not Working?

  #33 opened 5 years ago by rpand002
  4

• It seems that there lost a parentheses in the compute_hessian function. -- (p-n) / 2.*eps

  #34 opened 5 years ago by Frizy-up
  1

• Error in deriving the final genotype

  #38 opened 4 years ago by xiangzhluo
  0

• result not same as REAME.md

  #28 opened 5 years ago by YoungDav
  1

• nothing happens after running search.py

  #37 opened 5 years ago by yudabi
  1

• Does anyone meet the problem of memory leak? I trained this code on my own dataset and found that the memory keep rising during training, and I have no idea to solve it.

  #35 opened 5 years ago by ZuhaoLiu
  0

• validate loss cannot update

  #24 opened 5 years ago by Arui1
  2

• why darts only halved the image twice in the main part?

  #32 opened 5 years ago by JiyueWang
  0

• How to install graphviz?

  #31 opened 5 years ago by JiyueWang
  1

• Is there anyone who can reproduce the result ?

  #30 opened 5 years ago by heartInsert
  0

• Low CIFAR-100 accuracy

  #15 opened 5 years ago by 0xsamgreen
  4

• questions on v_grads = torch.autograd.grad(loss,v_alphas+v_weights)

  #27 opened 5 years ago by cholihao
  0

• Search runs only for 1 iteration

  #22 opened 5 years ago by IlyaTrofimov
  7

• Is it proper to use valid_loader in both train and validation?

  #25 opened 5 years ago by kamata1729
  2

• Running via Google Colab

  #26 opened 5 years ago by suryakumara
  0

• Suggestion for torch.backends.cudnn.benchmark = True in search process

  #23 opened 5 years ago by marsggbo
  0

• Could you please tell me how do you get the random search baseline

  #21 opened 5 years ago by leoozy
  0

• what's the use of augment.py

  #19 opened 5 years ago by johsnows
  2

• How to train on imagenet?THX.

  #14 opened 5 years ago by betterhalfwzm
  1

• question about SearchCNN code

  #20 opened 5 years ago by Christian-lyc
  0

• featuremap size reduce in op.py FactorizedReduce function

  #18 opened 5 years ago by yjump
  0

• meaning of Normal and Reduce?

  #17 opened 5 years ago by Damax18
  0

• Missing Softmax for Genotype selection

  #16 opened 5 years ago
  3

• TypeError: __init__() got an unexpected keyword argument 'log_dir'

  #13 opened 5 years ago by zjplab
  1

• How to use multi-gpu

  #2 opened 6 years ago by VectorYoung
  5

• Why is the search accuracy on the CIFAR-10 dataset only 89% accuracy?

  #10 opened 6 years ago by leidaguo
  2

• Augment

  #5 opened 5 years ago by TianMingChen
  5

• Cannot obtain the cell architecture as the reported one.

  #9 opened 5 years ago by fanglinpu
  1

• Why using broadcast for edge weights?

  #12 opened 5 years ago by zzzxxxttt
  2

• AssertionError: can only join a child process

  #8 opened 6 years ago by twangnh
  0

• License

  #7 opened 6 years ago by DaStoll
  2

• How to run the Augment

  #1 opened 6 years ago by VectorYoung
  1

• Where do I control the connections between nodes?

  #6 opened 6 years ago by sunil3590
  1