Alexander-H-Liu/End-to-end-ASR-Pytorch

Issues

• validation wer of training mode and test mode are different.

  #64 opened 4 years ago by burin-n
  3

• Is the Pure CTC training tested?

  #81 opened 9 months ago by yifan-bao
  0

• Has someone successfully run the generate_vocab_file.py

  #80 opened 2 years ago by didi222-lqq
  1

• Is confrontation used in this code? As stated in the paper

  #77 opened 3 years ago by Possi12
  1

• If anyone trained this models,Please post librispeech WER and/or timit PER.

  #38 opened 5 years ago by shiva1393
  1

• how to change GPU number

  #78 opened 3 years ago by diveknown
  1

• 推理过程的CTC打分

  #76 opened 3 years ago by clay-nuyoah
  0

• What is the ctc loss? The model is considered to be successfully trained

  #75 opened 3 years ago by scuizhibin
  1

• Has anyone successfully run this code on a Windows laptop?

  #73 opened 4 years ago by Zyhjoker
  1

• How to apply this repo to non-English language?

  #74 opened 3 years ago by supernirmesh
  1

• words alignment

  #72 opened 4 years ago by dangvansam
  0

• Hi, what is the longest audio for your work？

  #71 opened 4 years ago by Liquor520
  0

• any body test in aishell1?

  #70 opened 4 years ago by dayu1979
  1

• can I perform the experiment on WINDOWS?

  #56 opened 5 years ago by jiexiabalaa
  1

• Step 0 : text encoding

  #60 opened 4 years ago by taylerpauls
  1

• pretrained model

  #40 opened 5 years ago by smallflyingpig
  1

• Not support mfcc?

  #68 opened 4 years ago by ChangNamAn
  0

• Inference is extremely slow

  #30 opened 6 years ago by kamilkk852
  5

• what is the "subword" mean

  #66 opened 4 years ago by xjwla
  0

• Is SpecAugment supported?

  #65 opened 4 years ago by kouohhashi
  0

• TypeError: create_dataset() missing 1 required positional argument: 'test_path'

  #63 opened 4 years ago by Deep-learning999
  1

• 中英夾雜模型

  #59 opened 5 years ago by sungyihsun
  0

• Config file for Librispeech 960h

  #37 opened 5 years ago by Youyoun
  2

• predict have a problem

  #58 opened 5 years ago by jiashihuigithub
  0

• Single file Inference

  #54 opened 5 years ago by shamil-kadavan
  2

• Is there anybody uses custom data using librosa for traning LAS

  #55 opened 5 years ago by dobby-seo
  0

• Facing missing positional argument : 'init_adadelta'

  #53 opened 5 years ago by shamil-kadavan
  3

• Error in CharacterTextEncoder

  #51 opened 5 years ago by gheyret
  1

• Is it a bug on line 88 in preprocess.py

  #43 opened 5 years ago by jackaduma
  1

• Single audio inference

  #33 opened 5 years ago by iamxiaoyubei
  3

• What's the WER if training on full librispeech train set?

  #35 opened 5 years ago by iamxiaoyubei
  5

• libgomp: Thread creation failed: Resource temporarily unavailable

  #50 opened 5 years ago by nobel861017
  1

• Multihead attention

  #44 opened 5 years ago by d223302
  1

• Question about speller hidden initialization

  #26 opened 5 years ago by Andras7
  1

• Multi-GPU implementation

  #23 opened 5 years ago by kfmn
  1

• Padding target strings with the SOS token

  #22 opened 5 years ago by atliSig
  2

• Very different loss on validation

  #31 opened 6 years ago by kamilkk852
  1

• Segmentation fault (core dumped) when training on TIMIT

  #32 opened 5 years ago by iamxiaoyubei
  1

• Is this repo still maintained ?

  #36 opened 5 years ago by Youyoun
  1

• no text encoder training in timit preprocess

  #41 opened 5 years ago by smallflyingpig
  0

• error running example code line

  #39 opened 5 years ago by MNCTTY
  0

• I met Problem when running preprocess_libri.py

  #28 opened 6 years ago by qycxzhangyi
  15

• Could you mind tell me the WER of librispeech datasets? thanks

  #27 opened 5 years ago by iamxiaoyubei
  0

• FileNotFoundError: [Errno 2] No such file or directory: 'result/libri_example_sd0/asr'

  #29 opened 6 years ago by Mohammadelc
  0

• FileNotFoundError: [Errno 2] No such file or directory: 'spm_train'

  #25 opened 6 years ago by YasinZhao
  4

• Encoding Target

  #18 opened 6 years ago by MishaimMalik
  1

• About loading pretrained models

  #24 opened 6 years ago by atliSig
  0

• AttributeError: module 'torch.nn' has no attribute 'CTCLoss'

  #20 opened 6 years ago by profsatwinder
  1

• which state vector of multiple Speller layers is suggested to be used for attention calculation

  #21 opened 6 years ago by XuesongYang
  2

• Questions regarding ctc prefix score algorithm

  #16 opened 6 years ago by Chung-I
  1