automan000/Convolutional_LSTM_PyTorch

Issues

• Why self.num_features=4 in line 15?

  #21 opened 4 years ago by LeronQ
  0

• Input shape issue and lack of bias.

  #15 opened 5 years ago by mikumeow
  9

• Why Wci,Wcf,Wco should be initialized at the beginning of each batch?

  #9 opened 5 years ago by SampsonKwan
  6

• 关于input 和 step 的问题

  #7 opened 6 years ago by Fzz123
  2

• Where is the squence length

  #18 opened 5 years ago by 2294462474
  1

• Peephole connections (Wci, Wcf, Wco) gradient update

  #20 opened 5 years ago by Pozimek
  1

• why hidden_channels % 2 == 0 ?

  #19 opened 5 years ago by Muggle-K
  0

• Why Wci, Wcf, Wco are Variables rather than nn.Parameters

  #17 opened 5 years ago by bkoyuncu
  3

• The shape of the output

  #16 opened 5 years ago by Hiker01
  0

• What's the shape of input?

  #12 opened 5 years ago by mikumeow
  1

• I think forward code is wrong

  #13 opened 5 years ago by caffeinism
  0

• Found the code is much slower than Keras counterpart (takes 2-3 times longer time). Do you know why?

  #11 opened 5 years ago by xychenunc
  0

• Is there any experiments results provided?

  #10 opened 5 years ago
  0

• How to do the entire sequence all at once?

  #8 opened 6 years ago by SampsonKwan
  0

• Error in backward

  #2 opened 6 years ago by xgniu
  1

• RuntimeError: Jacobian mismatch for output 0 with respect to input 0

  #6 opened 6 years ago by inkplay
  3

• about forward

  #5 opened 6 years ago by chencodeX
  1

• about concat

  #4 opened 6 years ago by chencodeX
  5

• How can I use this module for predicting the moving mnist like the paper?

  #3 opened 6 years ago by guanfuchen
  1

• using for custom dataset

  #1 opened 6 years ago by KyonP
  1