run_dnn_fbank.sh
4.69 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
#!/bin/bash
# Copyright 2012-2014 Brno University of Technology (Author: Karel Vesely)
# Copyright 2016 Radboud University (Author: Emre Yilmaz)
# Apache 2.0
# This example script trains a DNN on top of FBANK features.
# The training is done in 3 stages,
#
# 1) RBM pre-training:
# in this unsupervised stage we train stack of RBMs,
# a good starting point for frame cross-entropy trainig.
# 2) frame cross-entropy training:
# the objective is to classify frames to correct pdfs.
# 3) sequence-training optimizing sMBR:
# the objective is to emphasize state-sequences with better
# frame accuracy w.r.t. reference alignment.
# Note: With DNNs in RM, the optimal LMWT is 2-6. Don't be tempted to try acwt's like 0.2,
# the value 0.1 is better both for decoding and sMBR.
. ./cmd.sh ## You'll want to change cmd.sh to something that will work on your system.
## This relates to the queue.
. ./path.sh ## Source the tools/utils (import the queue.pl)
dev=data-fbank/devel
tst=data-fbank/test
train=data-fbank/train
dev_original=data/devel
tst_original=data/test
train_original=data/train
gmm=exp/tri3
stage=0
. utils/parse_options.sh || exit 1;
set -eu
# Make the FBANK features
[ ! -e $dev ] && if [ $stage -le 0 ]; then
# Dev set
utils/copy_data_dir.sh $dev_original $dev || exit 1; rm $dev/{cmvn,feats}.scp
steps/make_fbank.sh --nj 10 --cmd "$train_cmd" \
$dev $dev/log $dev/data || exit 1;
steps/compute_cmvn_stats.sh $dev $dev/log $dev/data || exit 1;
# Test set
utils/copy_data_dir.sh $tst_original $tst || exit 1; rm $tst/{cmvn,feats}.scp
steps/make_fbank.sh --nj 10 --cmd "$train_cmd" \
$tst $tst/log $tst/data || exit 1;
steps/compute_cmvn_stats.sh $tst $tst/log $tst/data || exit 1;
# Training set
utils/copy_data_dir.sh $train_original $train || exit 1; rm $train/{cmvn,feats}.scp
steps/make_fbank.sh --nj 10 --cmd "$train_cmd" \
$train $train/log $train/data || exit 1;
steps/compute_cmvn_stats.sh $train $train/log $train/data || exit 1;
# Split the training set
utils/subset_data_dir_tr_cv.sh --cv-spk-percent 10 $train ${train}_tr90 ${train}_cv10
fi
if [ $stage -le 1 ]; then
# Pre-train DBN, i.e. a stack of RBMs (small database, smaller DNN)
dir=exp/dnn4d-fbank_pretrain-dbn
$cuda_cmd $dir/log/pretrain_dbn.log \
steps/nnet/pretrain_dbn.sh \
--cmvn-opts "--norm-means=true --norm-vars=true" \
--delta-opts "--delta-order=2" --splice 5 \
--hid-dim 2048 --rbm-iter 10 $train $dir || exit 1;
fi
if [ $stage -le 2 ]; then
# Train the DNN optimizing per-frame cross-entropy.
dir=exp/dnn4d-fbank_pretrain-dbn_dnn
ali=${gmm}_ali
feature_transform=exp/dnn4d-fbank_pretrain-dbn/final.feature_transform
dbn=exp/dnn4d-fbank_pretrain-dbn/6.dbn
# Train
$cuda_cmd $dir/log/train_nnet.log \
steps/nnet/train.sh --feature-transform $feature_transform --dbn $dbn --hid-layers 0 --learn-rate 0.008 \
${train}_tr90 ${train}_cv10 data/lang $ali $ali $dir || exit 1;
# Decode (reuse HCLG graph)
steps/nnet/decode.sh --nj 20 --cmd "$decode_cmd" --config conf/decode_dnn.config --acwt 0.1 \
$gmm/graph $dev $dir/decode_devel || exit 1;
steps/nnet/decode.sh --nj 20 --cmd "$decode_cmd" --config conf/decode_dnn.config --acwt 0.1 \
$gmm/graph $tst $dir/decode_test || exit 1;
fi
# Sequence training using sMBR criterion, we do Stochastic-GD with per-utterance updates.
# Note: With DNNs in RM, the optimal LMWT is 2-6. Don't be tempted to try acwt's like 0.2,
# the value 0.1 is better both for decoding and sMBR.
dir=exp/dnn4d-fbank_pretrain-dbn_dnn_smbr
srcdir=exp/dnn4d-fbank_pretrain-dbn_dnn
acwt=0.1
if [ $stage -le 3 ]; then
# First we generate lattices and alignments:
steps/nnet/align.sh --nj 20 --cmd "$train_cmd" \
$train data/lang $srcdir ${srcdir}_ali || exit 1;
steps/nnet/make_denlats.sh --nj 20 --cmd "$decode_cmd" --config conf/decode_dnn.config --acwt $acwt \
$train data/lang $srcdir ${srcdir}_denlats || exit 1;
fi
if [ $stage -le 4 ]; then
# Re-train the DNN by 6 iterations of sMBR
steps/nnet/train_mpe.sh --cmd "$cuda_cmd" --num-iters 6 --acwt $acwt --do-smbr true \
$train data/lang $srcdir ${srcdir}_ali ${srcdir}_denlats $dir || exit 1
# Decode
for ITER in 6 3 1; do
steps/nnet/decode.sh --nj 20 --cmd "$decode_cmd" --config conf/decode_dnn.config \
--nnet $dir/${ITER}.nnet --acwt $acwt \
$gmm/graph $dev $dir/decode_devel_it${ITER} || exit 1
steps/nnet/decode.sh --nj 20 --cmd "$decode_cmd" --config conf/decode_dnn.config \
--nnet $dir/${ITER}.nnet --acwt $acwt \
$gmm/graph $tst $dir/decode_test_it${ITER} || exit 1
done
fi
echo Success
exit 0
# Getting results [see RESULTS file]
# for x in exp/*/decode*; do [ -d $x ] && grep WER $x/wer_* | utils/best_wer.sh; done