mravanelli/pytorch-kaldi

Issues

• res.res

  #259 opened 2 years ago by LYPinASR
  0

• using different features instead of FMLLR

  #257 opened 3 years ago by Miamoto
  0

• ValueError: all the input array dimensions for the concatenation axis must match exactly, but along dimension 0, the array at index 0 has size 5658 and the array at index 1 has size 5640

  #234 opened 5 years ago by aziryasin
  2

• err_te is 1

  #256 opened 3 years ago by severusbunny
  0

• Use final_architecture1.pkl for live test

  #255 opened 3 years ago by spacecd
  4

• Before switch to SpeechBrain, how to use trained model in pytorch

  #254 opened 4 years ago by sun-peach
  0

• Can we resume training from the epoch we got interruption

  #241 opened 4 years ago by sun-peach
  4

• Unable to run forwarding step on test set

  #253 opened 4 years ago by kevinmchu
  0

• x-vector DNN model

  #252 opened 4 years ago by jvel07
  0

• How to setup parameters in "cfg/TIMIT_baselines/TIMIT_liGRU_fmllr.cfg"?

  #239 opened 4 years ago by ReinholdM
  1

• How to train/decode on reverberant speech?

  #251 opened 4 years ago by kevinmchu
  1

• No Decoding Output

  #250 opened 4 years ago by kevinmchu
  20

• Does pytorch-kaldi support chain model training?

  #248 opened 4 years ago by Vanka0051
  1

• Support for torch.nn.Transformer Class?

  #245 opened 4 years ago by niticon
  1

• Question about the Dimension of wx.0.weight in my mlp model

  #243 opened 4 years ago by zyn2609530
  1

• Word transcription of TIMIT dataset

  #249 opened 4 years ago by shessam
  1

• No WER stdout when decoding

  #247 opened 4 years ago by hellboywyh
  0

• The loss curve of train and dev is reasonable but why the Test Error keeps 53% or so?

  #244 opened 4 years ago by hellboywyh
  8

• KaldiFatalError during decoding phase

  #246 opened 4 years ago by tanzyy96
  0

• input shape of nns

  #242 opened 4 years ago by HaoranWeiUTD
  3

• Do bidirectional layers share the input-to-hidden weights?

  #240 opened 4 years ago by timolohrenz
  2

• How can I load a trained model in python with torch?

  #236 opened 4 years ago by sun-peach
  2

• Training on multi-gpu very slow

  #237 opened 4 years ago by sun-peach
  4

• How to setup parameters in "cfg/TIMIT_baselines/TIMIT_liGRU_fmllr.cfg"?

  #238 opened 4 years ago by ReinholdM
  6

• bad forward .ark file output when out model is a sequential model

  #235 opened 4 years ago by timolohrenz
  0

• Issue:IndexError: list index out of range

  #203 opened 5 years ago by chijiudao
  1

• Librispeech: Adding lattice rescoring

  #232 opened 5 years ago by omprakashsonie
  12

• Use arch_pretrain_file and got cuda runtime error

  #206 opened 5 years ago by gezhaoDL
  1

• How to load my pretrained model.

  #207 opened 5 years ago by couragelfyang
  2

• Error in labels / final.mdl

  #208 opened 5 years ago by bsutapa
  1

• Cannot get WER when running TIMIT_SincNet_raw experiment

  #211 opened 5 years ago by zehaitu
  2

• Cannot reproduce LSTM result on TIMIT

  #216 opened 5 years ago by arvoelke
  7

• Loss not decreasing for Hybrid CNN+DNN and CNN+BLSTM models.

  #223 opened 5 years ago by bipashasen
  1

• Chunk Mean and Variance Normalization

  #233 opened 5 years ago by timolohrenz
  3

• Error in the decoding

  #218 opened 5 years ago by mnabihali
  11

• Problem with loading rawwav as feature, it costs too much time

  #212 opened 5 years ago by liuzz13
  2

• Is there a problem with bidirectional LSTM？

  #214 opened 5 years ago by lezasantaizi
  1

• The parameter update of sincconv

  #224 opened 5 years ago by piplyman
  1

• Production mode save_outputs: getting bad .ark files

  #228 opened 5 years ago by matthewkperez
  1

• Does mini batch order matter?

  #219 opened 5 years ago by matthewkperez
  4

• Support more models of kaldi like chain model

  #222 opened 5 years ago by a550461053
  2

• Silence detection (VAD)

  #221 opened 5 years ago by matthewkperez
  1

• Error with getting shared_list

  #220 opened 5 years ago by seas2nada
  0

• ERROR: hmm-info command doesn't exist. However I have `hmm-info` working in terminal.

  #217 opened 5 years ago by andi611
  1

• ValueError: Precision not allowed in integer format specifier

  #215 opened 5 years ago by mnabihali
  13

• How could I get the complete decoding result instead of PARTIAL result when decoding.

  #213 opened 5 years ago by xixihahaggg
  1

• Training with multiple features

  #209 opened 5 years ago by abnerLing
  0

• Online speech recognition

  #210 opened 5 years ago by victornoriega
  5

• Is there Conv2d CNN demo for neural_network.py ?

  #205 opened 5 years ago by lezasantaizi
  2

• Can I continue the training after 24 epochs finished，how to do？

  #204 opened 5 years ago by gezhaoDL
  0