wav 轉 kaldi data 問題

[leo424y]

我已根據文件準備檔案，但碰到error

• fix_data_dir.sh: no utterances remained: not proceeding further." 相關源碼在

https://github.com/kaldi-asr/kaldi/blob/e31776f8e2a9c403c8c02993b795a61cc5067450/egs/wsj/s5/utils/fix_data_dir.sh#L154-L165

應是我5個資料有沒設對的地方，內容如下

請教隊長該如何解？

- data
  - text
    "w02001-A_000000-005800"
  - segments
    "sw02001-A_000000-005800 sw02001-A 0.000 58.000"
  - utt2spk
    "w02001-A_000000-005800 2001-A"
  - reco2file_and_channel
    "sw02001-A sw02001 A"
  - wav.scp
    "sw02001-A sox -G /usr/local/kaldi/egs/taiwanese/s5c/exp/model/decode_hok8bu7_1/r.wav -b 16 -c 1 -r 16k -t wav - |"

- error
  "# bash 服務來試.sh exp/tri4/graph data/train exp/model/decode_hok8bu7_1
  fix_data_dir.sh: no utterances remained: not proceeding further."

-  略掉 fix_data_dir.sh
  "root@f458340bf683:/usr/local/kaldi/egs/taiwanese/s5c# bash fw.sh exp/tri1/graph data/train exp/model/decode_hok8bu7_1
  steps/make_mfcc.sh --nj 1 --cmd run.pl exp/model/decode_hok8bu7_1 exp/model/decode_hok8bu7_1/make_mfcc/ exp/model/decode_hok8bu7_1/mfcc
  utils/validate_data_dir.sh: WARNING: you have only one speaker.  This probably a bad idea.
     Search for the word 'bold' in http://kaldi-asr.org/doc/data_prep.html
     for more information.
  utils/validate_data_dir.sh: Utterance list differs between exp/model/decode_hok8bu7_1/utt2spk and exp/model/decode_hok8bu7_1/segments
  utils/validate_data_dir.sh: Lengths are 1 vs 1"

ref

# This script makes sure that only the segments present in
# all of "feats.scp", "wav.scp" [if present], segments [if present]
# text, and utt2spk are present in any of them.
# It puts the original contents of data-dir into
# data-dir/.backup

https://github.com/kaldi-asr/kaldi/blob/master/egs/wsj/s5/utils/fix_data_dir.sh

[sih4sing5hong5]

有的號做w02001，有的號做sw02001，對--袂-起-來

[leo424y]

哎呀！我眼花了，謝謝隊長神救援！！