uoguelph-mlrg/theano_alexnet

Issues

• ValueError: GpuElemwise. Input dimension mis-match

  #39 opened 4 years ago by lidaiqiang-vsg
  0

• version of dependencies

  #38 opened 7 years ago by G33kyKitty
  1

• Execute alexnet using ligbpuarray backend

  #31 opened 8 years ago by deepali-c
  8

• latest update to conda and I started getting this issue: TypeError: slice indices must be integers or None or have an __index__ method

  #37 opened 7 years ago by Magotraa
  2

• Preprocessing with respect to aspect ratio

  #36 opened 7 years ago by hma02
  0

• An update must have the same type as the original shared variable

  #35 opened 7 years ago by DennisBang
  1

• error on Windows 10

  #33 opened 8 years ago by Magotraa
  22

• For new dataset

  #34 opened 7 years ago by DennisBang
  1

• Training Cost NAN

  #27 opened 9 years ago by jiangqy
  11

• pycuda mem_get_ipc_handle() error on Windows 10

  #32 opened 8 years ago by goffredogiordano
  6

• Merge split layer weights

  #30 opened 8 years ago by npit
  3

• Error with theano v0.9.0

  #29 opened 8 years ago by deepali-c
  3

• Forward and Backward Propagation

  #28 opened 9 years ago by ksarker1205
  1

• Validate Performance

  #25 opened 9 years ago by ksarker1205
  1

• Step_idx bug when resuming training

  #26 opened 9 years ago by hma02
  0

• Momentum loading

  #22 opened 9 years ago by am-gill
  4

• Testing network on CPU

  #23 opened 9 years ago by nguyenminhduc9988
  4

• meta data files

  #21 opened 9 years ago by srp1970
  1

• crop size and difference between cudnn and cuda-convnet

  #20 opened 9 years ago by soheilb
  4

• validation data preparation wrong in get_val_error_loss

  #6 opened 9 years ago by yanshuaicao
  4

• Other datasets.

  #7 opened 9 years ago by gasiort
  3

• Mean image for the pre-trained net?

  #15 opened 9 years ago by murbard
  5

• can't reproduce results with pre-trained models

  #19 opened 9 years ago by mesnilgr
  5

• Can you also use theano_alexnet and load weights from the caffe trained network?

  #17 opened 9 years ago by baptist
  0

• what is the dynamic range of the input pixels that the network accepts?

  #10 opened 9 years ago by oogilyboogily
  1

• Results from intermediate stages from ConvPoolLayer

  #9 opened 9 years ago by tumble-weed
  3

• Will train.py end itself?

  #8 opened 10 years ago by duguyue100
  5

• Trained model file

  #3 opened 10 years ago by erfannoury
  5

• There may be something wrong with your code:

  #2 opened 10 years ago by jingruixiaozhuang
  1

• RuntimeError "Could not compile cuda_convnet"

  #1 opened 10 years ago by ibogun
  2