Menu ▾ ▴

The problem of online-audio-server-decode-faster

Help
jesjkl
2014-06-16
2014-06-25

  • jesjkl

    jesjkl - 2014-06-16

    Hi,
    I tried to start service online-audio-server-decode-faster with parameter like this

    online-audio-server-decode-faster --verbose=1 --rt-min=0.5 --rt-max=3.0 --max-active=6000 --beam=72.0 --acoustic-scale=0.0769 final.mdl graph_test_fg/HCLG.fst graph_test_fg/words.txt '1:2:3:4:5' graph_test_fg/phones/word_boundary.int 5010 final.mat

    i got the right message：
    TcpServer: Listening on port: 5010
    Reading LDA matrix: final.mat...
    Reading acoustic model: final.mdl...
    Reading word list: graph_test_fg/words.txt...
    Reading word boundary file: graph_test_fg/phones/word_boundary.int...
    Reading FST: graph_test_fg/HCLG.fst...
    Waiting for client...

    and then i tried to connect with online-audio-client with this parameter
    online-audio-client --htk --vtt localhost 5010 scp:test.scp

    and my test.scp file content is：
    test.wav

    i got a error message from server
    ERROR (online-audio-server-decode-faster:OnlineLdaInput():online-feat-input.cc:209) Invalid parameters supplied to OnlineLdaInput
    ERROR (online-audio-server-decode-faster:OnlineLdaInput():online-feat-input.cc:209) Invalid parameters supplied to OnlineLdaInput

    [stack trace: ]
    kaldi::KaldiGetStackTrace()
    kaldi::KaldiErrorMessage::~KaldiErrorMessage()
    kaldi::OnlineLdaInput::OnlineLdaInput(kaldi::OnlineFeatInputItf*, kaldi::Matrix<float> const&, int, int)
    online-audio-server-decode-faster(main+0xe8b) [0x448ccf]
    /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed) [0x7f005288876d]
    online-audio-server-decode-faster() [0x447d89]</float>

    How can I solve it?
    Thanks.

     

    • Daniel Povey

      Daniel Povey - 2014-06-16

      You may have to set the --left-context and --right-context options to the
      same values used in training. (e.g. 3).
      Dan

      On Mon, Jun 16, 2014 at 5:14 AM, jesjkl jesjkl@users.sf.net wrote:

      Hi,
      I tried to start service online-audio-server-decode-faster with parameter
      like this

      online-audio-server-decode-faster --verbose=1 --rt-min=0.5 --rt-max=3.0
      --max-active=6000 --beam=72.0 --acoustic-scale=0.0769 final.mdl
      graph_test_fg/HCLG.fst graph_test_fg/words.txt '1:2:3:4:5'
      graph_test_fg/phones/word_boundary.int 5010 final.mat

      i got the right message：
      TcpServer: Listening on port: 5010
      Reading LDA matrix: final.mat...
      Reading acoustic model: final.mdl...
      Reading word list: graph_test_fg/words.txt...
      Reading word boundary file: graph_test_fg/phones/word_boundary.int...
      Reading FST: graph_test_fg/HCLG.fst...
      Waiting for client...

      and then i tried to connect with online-audio-client with this parameter
      online-audio-client --htk --vtt localhost 5010 scp:test.scp

      and my test.scp file content is：
      test.wav

      i got a error message from server
      ERROR
      (online-audio-server-decode-faster:OnlineLdaInput():online-feat-input.cc:209)
      Invalid parameters supplied to OnlineLdaInput
      ERROR
      (online-audio-server-decode-faster:OnlineLdaInput():online-feat-input.cc:209)
      Invalid parameters supplied to OnlineLdaInput

      [stack trace: ]
      kaldi::KaldiGetStackTrace()
      kaldi::KaldiErrorMessage::~KaldiErrorMessage()
      kaldi::OnlineLdaInput::OnlineLdaInput(kaldi::OnlineFeatInputItf*,
      kaldi::Matrix const&, int, int)
      online-audio-server-decode-faster(main+0xe8b) [0x448ccf]
      /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed) [0x7f005288876d]
      online-audio-server-decode-faster() [0x447d89]

      How can I solve it?
      Thanks.

      The problem of online-audio-server-decode-faster
      https://sourceforge.net/p/kaldi/discussion/1355348/thread/6dfefc73/?limit=25#b42e

      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/kaldi/discussion/1355348/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       

  • jesjkl

    jesjkl - 2014-06-17

    Thanks for Dan.

    i try to train command such as；
    steps/train_lda_mllt.sh --cmd "$train_cmd" --splice-opts "--left-context=3 --right-context=3" 8000 80000 data/train data/lang jesjkl/m2_ali jesjkl/m3
    and i set the --left-context=3 and --right-context=3 option, but i got the same error message.

    So, How can I solve it? Thanks.

     

    • Daniel Povey

      Daniel Povey - 2014-06-17

      No, I was saying you should give those options to
      online-audio-server-decode-faster.
      If that doesn't work, get it in a debugger and see the dimensions that it
      is comparing, that may give you a clue what has gone wrong.
      Dan

      On Mon, Jun 16, 2014 at 10:47 PM, jesjkl jesjkl@users.sf.net wrote:

      Thanks for Dan.

      i try to train command such as；
      steps/train_lda_mllt.sh --cmd "$train_cmd" --splice-opts "--left-context=3
      --right-context=3" 8000 80000 data/train data/lang jesjkl/m2_ali jesjkl/m3
      and i set the --left-context=3 and --right-context=3 option, but i got the
      same error message.

      So, How can I solve it? Thanks.

      The problem of online-audio-server-decode-faster
      https://sourceforge.net/p/kaldi/discussion/1355348/thread/6dfefc73/?limit=25#cac8

      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/kaldi/discussion/1355348/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       

  • jesjkl

    jesjkl - 2014-06-25

    Thanks for Dan.

    I have solve it.

     
MongoDB Logo MongoDB