nicodjimenez/lstm

Issues

• The backpropagation part is needed?

  #50 opened 3 years ago by xuqingyu26
  1

• You forget the tanh function in the last computation in the part of def bottom_data_is():

  #47 opened 5 years ago by mikechen66
  4

• why there is no 0.5* in the equation of the derivative of sate(t) ?

  #33 opened 3 years ago by tinymindxx
  4

• Error in using 2 outputs

  #48 opened 4 years ago by raj1052
  0

• The tanh_derivative should be : 1. + values**2

  #41 opened 6 years ago by Ningshiqi
  4

• about tan(s)

  #9 opened 8 years ago by lcdevelop
  4

• y_list dimension

  #1 opened 9 years ago by evbo
  7

• loss compute

  #46 opened 5 years ago by dwwx
  0

• lstm.py the 97 line

  #43 opened 5 years ago by chenyingjie2012
  0

• Explain self.wi = rand_arr(-0.1, 0.1, mem_cell_ct, concat_len)

  #27 opened 7 years ago by shaifugpt
  0

• self.state.h = self.state.s * self.state.o

  #42 opened 6 years ago by niumeng07
  1

• I don't understand wi_diff wf_diff etc

  #40 opened 6 years ago by kaikai-sk
  0

• I have a question of ds,how to compulete ds?

  #39 opened 6 years ago by kaikai-sk
  0

• can't convergence

  #37 opened 6 years ago by fmscole
  0

• new dataset

  #29 opened 7 years ago by KaaiLee
  1

• Execution problem

  #36 opened 6 years ago by vijayaprabakarank
  1

• How to predict a new sequence ?

  #34 opened 7 years ago by demonSong
  0

• Swift port

  #31 opened 7 years ago by emilianbold
  0

• mem_cell_ct =100 but only 4 LSTM nodes

  #12 opened 8 years ago by ichraibi
  5

• line 85&86:if s_prev == None: s_prev = np.zeros_like(self.state.s)

  #28 opened 7 years ago by dingqunfei
  5

• A parameter (x_dim) appears to be unused

  #26 opened 7 years ago by ruffrey
  0

• action recogniton

  #23 opened 7 years ago by Autoom
  0

• Usage of the LSTM

  #21 opened 8 years ago by robinjungers
  0

• Each cell gets x_input

  #20 opened 8 years ago by Demaunt
  0

• Schematic diagram of the Architecture?

  #16 opened 8 years ago by greninja
  0

• What does "M" really mean?

  #11 opened 8 years ago by xiaohu2015
  0

• Ported to Julia

  #10 opened 8 years ago by ehknight
  1

• No tanh on state.s in 'bottom_data_is'

  #7 opened 8 years ago by arryon
  2

• question

  #8 opened 8 years ago by lcdevelop
  2

• Port to D

  #6 opened 8 years ago by Marenz
  0

• After training, how to input some test samples?

  #5 opened 8 years ago by huangshichen
  1

• Where does constant error carousel come from?

  #2 opened 9 years ago by evbo
  7