liuzhuang13/slimming

Issues

• Channel number to be pruned must be a pow of 2 ?

  #12 opened 3 years ago by summer-wind-hi
  1

• sparse-training gets contrary BN distribution

  #11 opened 3 years ago by summer-wind-hi
  5

• Is it reasonable to get a threshold for all bn layers?

  #10 opened 5 years ago by zihaozhang9
  0

• Slimming Resnet

  #2 opened 7 years ago by hiyijian
  10

• cifar10 flops higher than cifar100 on DenseNet(40% pruned)

  #8 opened 5 years ago by Sirius083
  1

• 在剪枝程序vggprune.py中遇见程序报错

  #9 opened 5 years ago by xuemo188
  0

• What's the formulation of calculating flops?

  #6 opened 6 years ago by gaow0007
  1

• Slimming MobileNet

  #7 opened 6 years ago by liuzhihui2046
  1

• Slimming DenseNet

  #3 opened 7 years ago by haithanhp
  7

• Bias masking in BN layers

  #5 opened 6 years ago by KengChiLiu
  1

• local subgradient = S*torch.sign(weight)

  #4 opened 7 years ago by MrLinNing
  3

• Connection between last conv layer and fc layer

  #1 opened 7 years ago by xiao1228
  2