tobyyouup/conv_seq2seq

Issues

• tensorflow.python.framework.errors_impl.FailedPreconditionError: /; Is a directory

  #23 opened 6 years ago by jxz542189
  2

• Shape not compatible error

  #27 opened 5 years ago by Fancy7777
  0

• tensorflow version?

  #26 opened 5 years ago by xiaotaw
  0

• list some problems suffered in this project

  #25 opened 5 years ago by liuliu8622
  0

• question about position embedding.

  #24 opened 6 years ago by Jingyilang
  1

• Result on ReadMe ?

  #1 opened 7 years ago by dylanee2
  6

• The residual part in the decoder was perhaps implemented wrong?

  #17 opened 7 years ago by eduOS
  4

• Invalid Input Pipeline class: ParallelTextInputPipelineFairseq

  #22 opened 6 years ago by ConnieTong
  0

• Solutions of TypeError: <lambda>() got an unexpected keyword argument 'momentum'

  #21 opened 6 years ago by flycsuu
  0

• What is this suppose to do?

  #20 opened 6 years ago by Achilles-96
  0

• Adding external features

  #7 opened 6 years ago by laiviet
  0

• GLU

  #19 opened 6 years ago by WhuRan
  0

• Dataset

  #18 opened 6 years ago by ahyuan
  0

• got two questions about the conv_decoder_fairseq.py

  #12 opened 7 years ago by winnechan
  2

• Why is the future information removed in every layer?

  #16 opened 7 years ago by eduOS
  4

• dataset

  #14 opened 7 years ago by SeekPoint
  1

• invalidArgumentError (see above for traceback): Assign requires shapes of both tensors to match. lhs shape= [37007] rhs shape= [12003]

  #15 opened 7 years ago by SeekPoint
  1

• weight normalization is probably implemented wrong

  #13 opened 7 years ago by jingjing-gong
  1

• Training loss goes down and then go up

  #6 opened 7 years ago by jingjing-gong
  9

• Log file of training process

  #11 opened 7 years ago by Maneeeshita
  0

• Can not run with tf-1.2.1

  #10 opened 7 years ago by jiaxuehu
  0

• online inference or predict

  #9 opened 7 years ago by jacklone
  2

• Why not use updated parent beam's embedding as next input when inferring?

  #8 opened 7 years ago by nkqiaolin
  1

• How the output is fed back to the decoder?

  #5 opened 7 years ago by harsh-agarwal
  0

• Never see the pooling layers!!!

  #4 opened 7 years ago by linkerr
  1

• Where to initialize hash tables in a new model?

  #3 opened 7 years ago by anglil
  2

• Inference

  #2 opened 7 years ago by kaix90
  1