alphacep--vosk-api
Nickolay Shmyrev
87 行
3.0 KiB
Bash
可执行文件
87 行
3.0 KiB
Bash
可执行文件
#!/bin/bash
|
|
|
|
set -euo pipefail
|
|
|
|
# This script is called from local/chain_${suffix}/run_tdnn.sh and
|
|
# local/chain_${suffix}/run_tdnn.sh (and may eventually be called by more
|
|
# scripts). It contains the common feature preparation and
|
|
# iVector-related parts of the script. See those scripts for examples
|
|
# of usage.
|
|
|
|
stage=0
|
|
train_set=train
|
|
gmm=tri3
|
|
suffix=""
|
|
|
|
. ./cmd.sh
|
|
. ./path.sh
|
|
. utils/parse_options.sh
|
|
|
|
gmm_dir=exp/${gmm}
|
|
ali_dir=exp/${gmm}_ali
|
|
|
|
if [ $stage -le 4 ]; then
|
|
echo "$0: computing a subset of data to train the diagonal UBM."
|
|
# We'll use about a quarter of the data.
|
|
mkdir -p exp/chain${suffix}/diag_ubm
|
|
temp_data_root=exp/chain${suffix}/diag_ubm
|
|
|
|
num_utts_total=$(wc -l <data/${train_set}/utt2spk)
|
|
num_utts=$[$num_utts_total/4]
|
|
utils/data/subset_data_dir.sh data/${train_set} \
|
|
$num_utts ${temp_data_root}/${train_set}_subset
|
|
|
|
echo "$0: computing a PCA transform from the data."
|
|
steps/online/nnet2/get_pca_transform.sh --cmd "$train_cmd" \
|
|
--splice-opts "--left-context=3 --right-context=3" \
|
|
--max-utts 10000 --subsample 2 \
|
|
${temp_data_root}/${train_set}_subset \
|
|
exp/chain${suffix}/pca_transform
|
|
|
|
echo "$0: training the diagonal UBM."
|
|
# Use 512 Gaussians in the UBM.
|
|
steps/online/nnet2/train_diag_ubm.sh --cmd "$train_cmd" --nj 10 \
|
|
--num-frames 700000 \
|
|
--num-threads 8 \
|
|
${temp_data_root}/${train_set}_subset 512 \
|
|
exp/chain${suffix}/pca_transform exp/chain${suffix}/diag_ubm
|
|
fi
|
|
|
|
if [ $stage -le 5 ]; then
|
|
# Train the iVector extractor. Use all of the speed-perturbed data since iVector extractors
|
|
# can be sensitive to the amount of data. The script defaults to an iVector dimension of
|
|
# 100.
|
|
echo "$0: training the iVector extractor"
|
|
steps/online/nnet2/train_ivector_extractor.sh --cmd "$train_cmd" --nj 2 \
|
|
--ivector-dim 40 \
|
|
data/${train_set} exp/chain${suffix}/diag_ubm \
|
|
exp/chain${suffix}/extractor || exit 1;
|
|
fi
|
|
|
|
|
|
if [ $stage -le 6 ]; then
|
|
# We extract iVectors on the speed-perturbed training data after combining
|
|
# short segments, which will be what we train the system on. With
|
|
# --utts-per-spk-max 2, the script pairs the utterances into twos, and treats
|
|
# each of these pairs as one speaker; this gives more diversity in iVectors..
|
|
# Note that these are extracted 'online'.
|
|
|
|
# note, we don't encode the 'max2' in the name of the ivectordir even though
|
|
# that's the data we extract the ivectors from, as it's still going to be
|
|
# valid for the non-'max2' data, the utterance list is the same.
|
|
|
|
ivectordir=exp/chain${suffix}/ivectors_${train_set}
|
|
|
|
# having a larger number of speakers is helpful for generalization, and to
|
|
# handle per-utterance decoding well (iVector starts at zero).
|
|
temp_data_root=${ivectordir}
|
|
utils/data/modify_speaker_info.sh --utts-per-spk-max 2 \
|
|
data/${train_set} ${temp_data_root}/${train_set}_max2
|
|
|
|
steps/online/nnet2/extract_ivectors_online.sh --cmd "$train_cmd" --nj 10 \
|
|
${temp_data_root}/${train_set}_max2 \
|
|
exp/chain${suffix}/extractor $ivectordir
|
|
fi
|
|
|
|
exit 0
|