amaas/stanford-ctc

Issues

• Is there any language model pretrained that can be used directly?

  #13 opened 2 years ago by kobenaxie
  0

• Explanation of a part of k-prefix beam search decoding?

  #12 opened 5 years ago by nemkin
  1

• Beam search decoding

  #14 opened 5 years ago by githubharald
  1

• Arabic language compatibility

  #16 opened 6 years ago by abd-mhainy
  0

• Prune alpha, beta computation range

  #15 opened 7 years ago by vineelpratap
  0

• Which decoder to use

  #11 opened 7 years ago by hhbyyh
  0

• CTC for Bi-LSTM and Uni-LSTM

  #10 opened 8 years ago by double22a
  0

• p(blank symbol) >> p(non-blank symbol) during NN-CTC training

  #3 opened 9 years ago by lifelongeek
  9

• Questions regarding feature preprocessing, number of CTC outputs

  #1 opened 9 years ago by lifelongeek
  11

• On data formats and the way to ge them for dataLoader.py

  #9 opened 9 years ago by konts6102
  0

• Formatting of Training Data Files: keys, alis, feat

  #2 opened 9 years ago by mschonwe
  2

• what's the logic of performing normalization each time step in ctc?

  #8 opened 9 years ago
  2

• Multi-Utterance Mini-Batch

  #7 opened 9 years ago by mschonwe
  2

• CUDAMatrix Attribute error

  #6 opened 9 years ago by madhavsund
  2

• Missing Module: joblib

  #5 opened 9 years ago by mschonwe
  2

• ImportError: No module named fastdecode

  #4 opened 9 years ago by mschonwe
  2