train_dnn.sh
7.11 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
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
#!/bin/bash
# This script is based on egs/fisher_english/s5/run.sh. It trains a
# multisplice time-delay neural network used in the DNN-based speaker
# recognition recipes.
# It's best to run the commands in this one by one.
. ./cmd.sh
. ./path.sh
mfccdir=`pwd`/mfcc
set -e
# the next command produces the data in local/train_all_asr
local/dnn/fisher_data_prep.sh /export/corpora3/LDC/LDC2004T19 /export/corpora3/LDC/LDC2005T19 \
/export/corpora3/LDC/LDC2004S13 /export/corpora3/LDC/LDC2005S13
# You could also try specifying the --calldata argument to this command as below.
# If specified, the script will use actual speaker personal identification
# numbers released with the dataset, i.e. real speaker IDs. Note: --calldata has
# to be the first argument of this script.
# local/fisher_data_prep.sh --calldata /export/corpora3/LDC/LDC2004T19 /export/corpora3/LDC/LDC2005T19 \
# /export/corpora3/LDC/LDC2004S13 /export/corpora3/LDC/LDC2005S13
# at BUT:
# local/fisher_data_prep.sh /mnt/matylda6/jhu09/qpovey/FISHER/LDC2005T19 /mnt/matylda2/data/FISHER/
local/dnn/fisher_prepare_dict.sh
utils/prepare_lang.sh data/local/dict "<unk>" data/local/lang data/lang
local/dnn/fisher_train_lms.sh
local/dnn/fisher_create_test_lang.sh
# Use the first 4k sentences as dev set. Note: when we trained the LM, we used
# the 1st 10k sentences as dev set, so the 1st 4k won't have been used in the
# LM training data. However, they will be in the lexicon, plus speakers
# may overlap, so it's still not quite equivalent to a test set.
utils/fix_data_dir.sh data/train_all_asr
steps/make_mfcc.sh --nj 40 --cmd "$train_cmd" --mfcc-config conf/mfcc_asr.conf \
data/train_all_asr exp/make_mfcc/train_all_asr $mfccdir || exit 1;
utils/fix_data_dir.sh data/train_all_asr
utils/validate_data_dir.sh data/train_all_asr
# The dev and test sets are each about 3.3 hours long. These are not carefully
# done; there may be some speaker overlap with each other and with the training
# set. Note: in our LM-training setup we excluded the first 10k utterances (they
# were used for tuning but not for training), so the LM was not (directly) trained
# on either the dev or test sets.
utils/subset_data_dir.sh --first data/train_all_asr 10000 data/dev_and_test_asr
utils/subset_data_dir.sh --first data/dev_and_test_asr 5000 data/dev_asr
utils/subset_data_dir.sh --last data/dev_and_test_asr 5000 data/test_asr
rm -r data/dev_and_test_asr
steps/compute_cmvn_stats.sh data/dev_asr exp/make_mfcc/dev_asr $mfccdir
steps/compute_cmvn_stats.sh data/test_asr exp/make_mfcc/test_asr $mfccdir
n=$[`cat data/train_all_asr/segments | wc -l` - 10000]
utils/subset_data_dir.sh --last data/train_all_asr $n data/train_asr
steps/compute_cmvn_stats.sh data/train_asr exp/make_mfcc/train_asr $mfccdir
# Now-- there are 1.6 million utterances, and we want to start the monophone training
# on relatively short utterances (easier to align), but not only the very shortest
# ones (mostly uh-huh). So take the 100k shortest ones, and then take 10k random
# utterances from those.
utils/subset_data_dir.sh --shortest data/train_asr 100000 data/train_asr_100kshort
utils/subset_data_dir.sh data/train_asr_100kshort 10000 data/train_asr_10k
local/dnn/remove_dup_utts.sh 100 data/train_asr_10k data/train_asr_10k_nodup
utils/subset_data_dir.sh --speakers data/train_asr 30000 data/train_asr_30k
utils/subset_data_dir.sh --speakers data/train_asr 100000 data/train_asr_100k
# The next commands are not necessary for the scripts to run, but increase
# efficiency of data access by putting the mfcc's of the subset
# in a contiguous place in a file.
( . ./path.sh;
# make sure mfccdir is defined as above..
cp data/train_asr_10k_nodup/feats.scp{,.bak}
copy-feats scp:data/train_asr_10k_nodup/feats.scp ark,scp:$mfccdir/kaldi_fish_10k_nodup.ark,$mfccdir/kaldi_fish_10k_nodup.scp \
&& cp $mfccdir/kaldi_fish_10k_nodup.scp data/train_asr_10k_nodup/feats.scp
)
( . ./path.sh;
# make sure mfccdir is defined as above..
cp data/train_asr_30k/feats.scp{,.bak}
copy-feats scp:data/train_asr_30k/feats.scp ark,scp:$mfccdir/kaldi_fish_30k.ark,$mfccdir/kaldi_fish_30k.scp \
&& cp $mfccdir/kaldi_fish_30k.scp data/train_asr_30k/feats.scp
)
( . ./path.sh;
# make sure mfccdir is defined as above..
cp data/train_asr_100k/feats.scp{,.bak}
copy-feats scp:data/train_asr_100k/feats.scp ark,scp:$mfccdir/kaldi_fish_100k.ark,$mfccdir/kaldi_fish_100k.scp \
&& cp $mfccdir/kaldi_fish_100k.scp data/train_asr_100k/feats.scp
)
steps/train_mono.sh --nj 10 --cmd "$train_cmd" \
data/train_asr_10k_nodup data/lang exp/mono0a
steps/align_si.sh --nj 30 --cmd "$train_cmd" \
data/train_asr_30k data/lang exp/mono0a exp/mono0a_ali || exit 1;
steps/train_deltas.sh --cmd "$train_cmd" \
2500 20000 data/train_asr_30k data/lang exp/mono0a_ali exp/tri1 || exit 1;
(utils/mkgraph.sh data/lang_test exp/tri1 exp/tri1/graph
steps/decode.sh --nj 25 --cmd "$decode_cmd" --config conf/decode.config \
exp/tri1/graph data/dev exp/tri1/decode_dev)&
steps/align_si.sh --nj 30 --cmd "$train_cmd" \
data/train_asr_30k data/lang exp/tri1 exp/tri1_ali || exit 1;
steps/train_deltas.sh --cmd "$train_cmd" \
2500 20000 data/train_asr_30k data/lang exp/tri1_ali exp/tri2 || exit 1;
(
utils/mkgraph.sh data/lang_test exp/tri2 exp/tri2/graph || exit 1;
steps/decode.sh --nj 25 --cmd "$decode_cmd" --config conf/decode.config \
exp/tri2/graph data/dev exp/tri2/decode_dev || exit 1;
)&
steps/align_si.sh --nj 30 --cmd "$train_cmd" \
data/train_asr_100k data/lang exp/tri2 exp/tri2_ali || exit 1;
# Train tri3a, which is LDA+MLLT, on 100k data.
steps/train_lda_mllt.sh --cmd "$train_cmd" \
--splice-opts "--left-context=3 --right-context=3" \
5000 40000 data/train_asr_100k data/lang exp/tri2_ali exp/tri3a || exit 1;
(
utils/mkgraph.sh data/lang_test exp/tri3a exp/tri3a/graph || exit 1;
steps/decode.sh --nj 25 --cmd "$decode_cmd" --config conf/decode.config \
exp/tri3a/graph data/dev exp/tri3a/decode_dev || exit 1;
)&
# Next we'll use fMLLR and train with SAT (i.e. on
# fMLLR features)
steps/align_fmllr.sh --nj 30 --cmd "$train_cmd" \
data/train_asr_100k data/lang exp/tri3a exp/tri3a_ali || exit 1;
steps/train_sat.sh --cmd "$train_cmd" \
5000 100000 data/train_asr_100k data/lang exp/tri3a_ali exp/tri4a || exit 1;
(
utils/mkgraph.sh data/lang_test exp/tri4a exp/tri4a/graph
steps/decode_fmllr.sh --nj 25 --cmd "$decode_cmd" --config conf/decode.config \
exp/tri4a/graph data/dev exp/tri4a/decode_dev
)&
steps/align_fmllr.sh --nj 30 --cmd "$train_cmd" \
data/train_asr data/lang exp/tri4a exp/tri4a_ali || exit 1;
steps/train_sat.sh --cmd "$train_cmd" \
7000 300000 data/train_asr data/lang exp/tri4a_ali exp/tri5a || exit 1;
(
utils/mkgraph.sh data/lang_test exp/tri5a exp/tri5a/graph
steps/decode_fmllr.sh --nj 25 --cmd "$decode_cmd" --config conf/decode.config \
exp/tri5a/graph data/dev exp/tri5a/decode_dev
)&
# this will help find issues with the lexicon.
# steps/cleanup/debug_lexicon.sh --nj 300 --cmd "$train_cmd" data/train_asr_100k data/lang exp/tri5a data/local/dict/lexicon.txt exp/debug_lexicon_100k
## The following is based on the best current neural net recipe.
local/dnn/run_nnet2_multisplice.sh