#!/bin/bash
  
  # Copyright 2016 Pegah Ghahremani
  
  # This scripts extracts iVector using global iVector extractor
  # trained on all languages in multilingual setup.
  
  . ./cmd.sh
  set -e
  stage=1
  train_set=train_sp_hires # train_set used to extract ivector using shared ivector
                           # extractor.
  ivector_suffix=_gb
  nnet3_affix=
  
  [ ! -f ./conf/common_vars.sh ] && echo 'the file conf/common_vars.sh does not exist!' && exit 1
  
  . conf/common_vars.sh || exit 1;
  
  . ./utils/parse_options.sh
  
  lang=$1
  global_extractor=$2
  
  if [ $stage -le 7 ]; then
    # We extract iVectors on all the train_nodup data, which will be what we
    # train the system on.
    # having a larger number of speakers is helpful for generalization, and to
    # handle per-utterance decoding well (iVector starts at zero).
    steps/online/nnet2/copy_data_dir.sh --utts-per-spk-max 2 data/$lang/${train_set} data/$lang/${train_set}_max2
    if [ ! -f exp/$lang/nnet3${nnet3_affix}/ivectors_${train_set}${ivector_suffix}/ivector_online.scp ]; then
      steps/online/nnet2/extract_ivectors_online.sh --cmd "$train_cmd" --nj 200 \
        data/$lang/${train_set}_max2 $global_extractor exp/$lang/nnet3${nnet3_affix}/ivectors_${train_set}${ivector_suffix} || exit 1;
    fi
  fi
  exit 0;