vosk-api
vosk-api copied to clipboard
I followed your steps for training, now how i get files to create model
i followed https://github.com/alphacep/vosk-api/tree/master/training
Please tell me which files i need to create model, please tell me steps.
i am also creating documentation for contribution.
please check if there is any error or not.
i just removed some same lines in output because github gives character limit error.
user@user:~/Documents/ML_Projects/vosk-api/kaldi/egs/wsj/training$ ./run.sh
Preparing phone lists and lexicon
utils/prepare_lang.sh data/local/dict <UNK> data/local/lang data/lang
Checking data/local/dict/silence_phones.txt ...
--> reading data/local/dict/silence_phones.txt
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/local/dict/silence_phones.txt is OK
Checking data/local/dict/optional_silence.txt ...
--> reading data/local/dict/optional_silence.txt
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/local/dict/optional_silence.txt is OK
Checking data/local/dict/nonsilence_phones.txt ...
--> reading data/local/dict/nonsilence_phones.txt
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/local/dict/nonsilence_phones.txt is OK
Checking disjoint: silence_phones.txt, nonsilence_phones.txt
--> disjoint property is OK.
Checking data/local/dict/lexicon.txt
--> reading data/local/dict/lexicon.txt
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/local/dict/lexicon.txt is OK
Checking data/local/dict/lexiconp.txt
--> reading data/local/dict/lexiconp.txt
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/local/dict/lexiconp.txt is OK
Checking lexicon pair data/local/dict/lexicon.txt and data/local/dict/lexiconp.txt
--> lexicon pair data/local/dict/lexicon.txt and data/local/dict/lexiconp.txt match
Checking data/local/dict/extra_questions.txt ...
--> data/local/dict/extra_questions.txt is empty (this is OK)
--> SUCCESS [validating dictionary directory data/local/dict]
fstaddselfloops data/lang/phones/wdisambig_phones.int data/lang/phones/wdisambig_words.int
prepare_lang.sh: validating output directory
utils/validate_lang.pl data/lang
Checking existence of separator file
separator file data/lang/subword_separator.txt is empty or does not exist, deal in word case.
Checking data/lang/phones.txt ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/lang/phones.txt is OK
Checking words.txt: #0 ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> data/lang/words.txt is OK
Checking disjoint: silence.txt, nonsilence.txt, disambig.txt ...
--> silence.txt and nonsilence.txt are disjoint
--> silence.txt and disambig.txt are disjoint
--> disambig.txt and nonsilence.txt are disjoint
--> disjoint property is OK
Checking sumation: silence.txt, nonsilence.txt, disambig.txt ...
--> found no unexplainable phones in phones.txt
Checking data/lang/phones/context_indep.{txt, int, csl} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 10 entry/entries in data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.int corresponds to data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.csl corresponds to data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.{txt, int, csl} are OK
Checking data/lang/phones/nonsilence.{txt, int, csl} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 156 entry/entries in data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.int corresponds to data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.csl corresponds to data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.{txt, int, csl} are OK
Checking data/lang/phones/silence.{txt, int, csl} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 10 entry/entries in data/lang/phones/silence.txt
--> data/lang/phones/silence.int corresponds to data/lang/phones/silence.txt
--> data/lang/phones/silence.csl corresponds to data/lang/phones/silence.txt
--> data/lang/phones/silence.{txt, int, csl} are OK
Checking data/lang/phones/optional_silence.{txt, int, csl} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 1 entry/entries in data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.int corresponds to data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.csl corresponds to data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.{txt, int, csl} are OK
Checking data/lang/phones/disambig.{txt, int, csl} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 21 entry/entries in data/lang/phones/disambig.txt
--> data/lang/phones/disambig.int corresponds to data/lang/phones/disambig.txt
--> data/lang/phones/disambig.csl corresponds to data/lang/phones/disambig.txt
--> data/lang/phones/disambig.{txt, int, csl} are OK
Checking data/lang/phones/roots.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 41 entry/entries in data/lang/phones/roots.txt
--> data/lang/phones/roots.int corresponds to data/lang/phones/roots.txt
--> data/lang/phones/roots.{txt, int} are OK
Checking data/lang/phones/sets.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 41 entry/entries in data/lang/phones/sets.txt
--> data/lang/phones/sets.int corresponds to data/lang/phones/sets.txt
--> data/lang/phones/sets.{txt, int} are OK
Checking data/lang/phones/extra_questions.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 9 entry/entries in data/lang/phones/extra_questions.txt
--> data/lang/phones/extra_questions.int corresponds to data/lang/phones/extra_questions.txt
--> data/lang/phones/extra_questions.{txt, int} are OK
Checking data/lang/phones/word_boundary.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 166 entry/entries in data/lang/phones/word_boundary.txt
--> data/lang/phones/word_boundary.int corresponds to data/lang/phones/word_boundary.txt
--> data/lang/phones/word_boundary.{txt, int} are OK
Checking optional_silence.txt ...
--> reading data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.txt is OK
Checking disambiguation symbols: #0 and #1
--> data/lang/phones/disambig.txt has "#0" and "#1"
--> data/lang/phones/disambig.txt is OK
Checking topo ...
Checking word_boundary.txt: silence.txt, nonsilence.txt, disambig.txt ...
--> data/lang/phones/word_boundary.txt doesn't include disambiguation symbols
--> data/lang/phones/word_boundary.txt is the union of nonsilence.txt and silence.txt
--> data/lang/phones/word_boundary.txt is OK
Checking word-level disambiguation symbols...
--> data/lang/phones/wdisambig.txt exists (newer prepare_lang.sh)
Checking word_boundary.int and disambig.int
--> generating a 79 word/subword sequence
--> resulting phone sequence from L.fst corresponds to the word sequence
--> L.fst is OK
--> generating a 16 word/subword sequence
--> resulting phone sequence from L_disambig.fst corresponds to the word sequence
--> L_disambig.fst is OK
Checking data/lang/oov.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 1 entry/entries in data/lang/oov.txt
--> data/lang/oov.int corresponds to data/lang/oov.txt
--> data/lang/oov.{txt, int} are OK
--> data/lang/L.fst is olabel sorted
--> data/lang/L_disambig.fst is olabel sorted
--> SUCCESS [validating lang directory data/lang]
steps/make_mfcc.sh --cmd run.pl --nj 10 data/train exp/make_mfcc/train
steps/make_mfcc.sh: moving data/train/feats.scp to data/train/.backup
utils/validate_data_dir.sh: Successfully validated data-directory data/train
steps/make_mfcc.sh: [info]: no segments file exists: assuming wav.scp indexed by utterance.
steps/make_mfcc.sh: Succeeded creating MFCC features for train
steps/compute_cmvn_stats.sh data/train exp/make_mfcc/train
Succeeded creating CMVN stats for train
steps/train_mono.sh --nj 10 --cmd run.pl data/train data/lang exp/mono
steps/train_mono.sh: Initializing monophone system.
steps/train_mono.sh: Compiling training graphs
steps/train_mono.sh: Aligning data equally (pass 0)
steps/train_mono.sh: Pass 1
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 2
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 3
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 4
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 5
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 6
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 7
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 8
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 9
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 10
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 11
steps/train_mono.sh: Pass 12
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 13
steps/train_mono.sh: Pass 14
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 15
steps/train_mono.sh: Pass 16
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 17
steps/train_mono.sh: Pass 18
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 19
steps/train_mono.sh: Pass 20
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 21
steps/train_mono.sh: Pass 22
steps/train_mono.sh: Pass 23
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 24
steps/train_mono.sh: Pass 25
steps/train_mono.sh: Pass 26
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 27
steps/train_mono.sh: Pass 28
steps/train_mono.sh: Pass 29
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 30
steps/train_mono.sh: Pass 31
steps/train_mono.sh: Pass 32
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 33
steps/train_mono.sh: Pass 34
steps/train_mono.sh: Pass 35
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 36
steps/train_mono.sh: Pass 37
steps/train_mono.sh: Pass 38
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 39
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/mono
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 43.3352810564% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 51.5087063223% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/mono/log/analyze_alignments.log
1741840 warnings in exp/mono/log/acc.*.*.log
3518420 warnings in exp/mono/log/align.*.*.log
2 warnings in exp/mono/log/analyze_alignments.log
exp/mono: nj=10 align prob=-243.19 over 91.90h [retry=36.7%, fail=14.4%] states=127 gauss=1001
steps/train_mono.sh: Done training monophone system in exp/mono
steps/align_si.sh --nj 10 --cmd run.pl data/train data/lang exp/mono exp/mono_ali
steps/align_si.sh: feature type is delta
steps/align_si.sh: aligning data in data/train using model from exp/mono, putting alignments in exp/mono_ali
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/mono_ali
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 43.1685571582% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 51.449376838% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/mono_ali/log/analyze_alignments.log
steps/align_si.sh: done aligning data.
steps/train_lda_mllt.sh --cmd run.pl 2000 10000 data/train data/lang exp/mono_ali exp/tri1
steps/train_lda_mllt.sh: Accumulating LDA statistics.
steps/train_lda_mllt.sh: Accumulating tree stats
steps/train_lda_mllt.sh: Getting questions for tree clustering.
steps/train_lda_mllt.sh: Building the tree
steps/train_lda_mllt.sh: Initializing the model
steps/train_lda_mllt.sh: Converting alignments from exp/mono_ali to use current tree
steps/train_lda_mllt.sh: Compiling graphs of transcripts
Training pass 1
Training pass 2
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 3
Training pass 4
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 5
Training pass 6
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 7
Training pass 8
Training pass 9
Training pass 10
Aligning data
Training pass 11
Training pass 12
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 13
Training pass 14
Training pass 15
Training pass 16
Training pass 17
Training pass 18
Training pass 19
Training pass 20
Aligning data
Training pass 21
Training pass 22
Training pass 23
Training pass 24
Training pass 25
Training pass 26
Training pass 27
Training pass 28
Training pass 29
Training pass 30
Aligning data
Training pass 31
Training pass 32
Training pass 33
Training pass 34
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri1
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 36.2239808472% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 46.7424796613% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri1/log/analyze_alignments.log
7143 warnings in exp/tri1/log/lda_acc.*.log
1638433 warnings in exp/tri1/log/acc.*.*.log
1 warnings in exp/tri1/log/update.*.log
2 warnings in exp/tri1/log/analyze_alignments.log
1 warnings in exp/tri1/log/build_tree.log
489716 warnings in exp/tri1/log/align.*.*.log
exp/tri1: nj=10 align prob=-48.04 over 96.59h [retry=27.1%, fail=9.8%] states=1680 gauss=10029 tree-impr=3.22 lda-sum=6.80 mllt:impr,logdet=0.77,1.07
steps/train_lda_mllt.sh: Done training system with LDA+MLLT features in exp/tri1
steps/align_si.sh --nj 10 --cmd run.pl data/train data/lang exp/tri1 exp/tri1_ali
steps/align_si.sh: feature type is lda
steps/align_si.sh: aligning data in data/train using model from exp/tri1, putting alignments in exp/tri1_ali
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri1_ali
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 36.016676646% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 46.5305036258% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri1_ali/log/analyze_alignments.log
steps/align_si.sh: done aligning data.
steps/train_lda_mllt.sh --cmd run.pl 2500 15000 data/train data/lang exp/tri1_ali exp/tri2
steps/train_lda_mllt.sh: Accumulating LDA statistics.
steps/train_lda_mllt.sh: Accumulating tree stats
steps/train_lda_mllt.sh: Getting questions for tree clustering.
steps/train_lda_mllt.sh: Building the tree
steps/train_lda_mllt.sh: Initializing the model
steps/train_lda_mllt.sh: Converting alignments from exp/tri1_ali to use current tree
steps/train_lda_mllt.sh: Compiling graphs of transcripts
Training pass 1
Training pass 2
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 3
Training pass 4
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 5
Training pass 6
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 7
Training pass 8
Training pass 9
Training pass 10
Aligning data
Training pass 11
Training pass 12
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 13
Training pass 14
Training pass 15
Training pass 16
Training pass 17
Training pass 18
Training pass 19
Training pass 20
Aligning data
Training pass 21
Training pass 22
Training pass 23
Training pass 24
Training pass 25
Training pass 26
Training pass 27
Training pass 28
Training pass 29
Training pass 30
Aligning data
Training pass 31
Training pass 32
Training pass 33
Training pass 34
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri2
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 35.4488086416% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 46.1002941903% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri2/log/analyze_alignments.log
4896 warnings in exp/tri2/log/lda_acc.*.log
1 warnings in exp/tri2/log/build_tree.log
57502 warnings in exp/tri2/log/align.*.*.log
181624 warnings in exp/tri2/log/acc.*.*.log
2 warnings in exp/tri2/log/analyze_alignments.log
exp/tri2: nj=10 align prob=-48.78 over 95.41h [retry=26.6%, fail=10.9%] states=2088 gauss=15028 tree-impr=4.16 lda-sum=15.78 mllt:impr,logdet=0.84,1.19
steps/train_lda_mllt.sh: Done training system with LDA+MLLT features in exp/tri2
steps/align_si.sh --nj 10 --cmd run.pl data/train data/lang exp/tri2 exp/tri2_ali
steps/align_si.sh: feature type is lda
steps/align_si.sh: aligning data in data/train using model from exp/tri2, putting alignments in exp/tri2_ali
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri2_ali
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 35.1991557771% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 45.817055099% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri2_ali/log/analyze_alignments.log
steps/align_si.sh: done aligning data.
steps/train_lda_mllt.sh --cmd run.pl 2500 20000 data/train data/lang exp/tri2_ali exp/tri3
steps/train_lda_mllt.sh: Accumulating LDA statistics.
steps/train_lda_mllt.sh: Accumulating tree stats
steps/train_lda_mllt.sh: Getting questions for tree clustering.
steps/train_lda_mllt.sh: Building the tree
steps/train_lda_mllt.sh: Initializing the model
steps/train_lda_mllt.sh: Converting alignments from exp/tri2_ali to use current tree
steps/train_lda_mllt.sh: Compiling graphs of transcripts
Training pass 1
Training pass 2
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 3
Training pass 4
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 5
Training pass 6
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 7
Training pass 8
Training pass 9
Training pass 10
Aligning data
Training pass 11
Training pass 12
steps/train_lda_mllt.sh: Estimating MLLT
Training pass 13
Training pass 14
Training pass 15
Training pass 16
Training pass 17
Training pass 18
Training pass 19
Training pass 20
Aligning data
Training pass 21
Training pass 22
Training pass 23
Training pass 24
Training pass 25
Training pass 26
Training pass 27
Training pass 28
Training pass 29
Training pass 30
Aligning data
Training pass 31
Training pass 32
Training pass 33
Training pass 34
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri3
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 33.1143853335% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 45.331233832% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri3/log/analyze_alignments.log
1 warnings in exp/tri3/log/build_tree.log
1 warnings in exp/tri3/log/update.*.log
191039 warnings in exp/tri3/log/acc.*.*.log
57639 warnings in exp/tri3/log/align.*.*.log
2 warnings in exp/tri3/log/analyze_alignments.log
5451 warnings in exp/tri3/log/lda_acc.*.log
exp/tri3: nj=10 align prob=-48.86 over 95.23h [retry=26.1%, fail=11.1%] states=2088 gauss=20029 tree-impr=4.34 lda-sum=17.19 mllt:impr,logdet=0.86,1.26
steps/train_lda_mllt.sh: Done training system with LDA+MLLT features in exp/tri3
steps/align_si.sh --nj 10 --cmd run.pl data/train data/lang exp/tri3 exp/tri3_ali
steps/align_si.sh: feature type is lda
steps/align_si.sh: aligning data in data/train using model from exp/tri3, putting alignments in exp/tri3_ali
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri3_ali
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 32.5647767366% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 45.1807905796% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri3_ali/log/analyze_alignments.log
steps/align_si.sh: done aligning data.
local/chain/run_tdnn.sh
local/chain/run_ivector_common.sh: computing a subset of data to train the diagonal UBM.
utils/data/subset_data_dir.sh: reducing #utt from 49989 to 12497
local/chain/run_ivector_common.sh: computing a PCA transform from the data.
steps/online/nnet2/get_pca_transform.sh --cmd run.pl --splice-opts --left-context=3 --right-context=3 --max-utts 10000 --subsample 2 exp/chain/diag_ubm/train_subset exp/chain/pca_transform
Done estimating PCA transform in exp/chain/pca_transform
local/chain/run_ivector_common.sh: training the diagonal UBM.
steps/online/nnet2/train_diag_ubm.sh --cmd run.pl --nj 10 --num-frames 700000 --num-threads 8 exp/chain/diag_ubm/train_subset 512 exp/chain/pca_transform exp/chain/diag_ubm
filter_scps.pl: warning: some input lines did not get output
steps/online/nnet2/train_diag_ubm.sh: Directory exp/chain/diag_ubm already exists. Backing up diagonal UBM in exp/chain/diag_ubm/backup.U4M
steps/online/nnet2/train_diag_ubm.sh: initializing model from E-M in memory,
steps/online/nnet2/train_diag_ubm.sh: starting from 256 Gaussians, reaching 512;
steps/online/nnet2/train_diag_ubm.sh: for 20 iterations, using at most 700000 frames of data
Getting Gaussian-selection info
steps/online/nnet2/train_diag_ubm.sh: will train for 4 iterations, in parallel over
steps/online/nnet2/train_diag_ubm.sh: 10 machines, parallelized with 'run.pl'
steps/online/nnet2/train_diag_ubm.sh: Training pass 0
steps/online/nnet2/train_diag_ubm.sh: Training pass 1
steps/online/nnet2/train_diag_ubm.sh: Training pass 2
steps/online/nnet2/train_diag_ubm.sh: Training pass 3
local/chain/run_ivector_common.sh: training the iVector extractor
steps/online/nnet2/train_ivector_extractor.sh --cmd run.pl --nj 2 --ivector-dim 40 data/train exp/chain/diag_ubm exp/chain/extractor
steps/online/nnet2/train_ivector_extractor.sh: Directory exp/chain/extractor already exists. Backing up iVector extractor in exp/chain/extractor/backup.O1v
steps/online/nnet2/train_ivector_extractor.sh: doing Gaussian selection and posterior computation
Accumulating stats (pass 0)
Summing accs (pass 0)
Updating model (pass 0)
Accumulating stats (pass 1)
Summing accs (pass 1)
Updating model (pass 1)
Accumulating stats (pass 2)
Summing accs (pass 2)
Updating model (pass 2)
Accumulating stats (pass 3)
Summing accs (pass 3)
Updating model (pass 3)
Accumulating stats (pass 4)
Summing accs (pass 4)
Updating model (pass 4)
Accumulating stats (pass 5)
Summing accs (pass 5)
Updating model (pass 5)
Accumulating stats (pass 6)
Summing accs (pass 6)
Updating model (pass 6)
Accumulating stats (pass 7)
Summing accs (pass 7)
Updating model (pass 7)
Accumulating stats (pass 8)
Summing accs (pass 8)
Updating model (pass 8)
Accumulating stats (pass 9)
Summing accs (pass 9)
Updating model (pass 9)
utils/data/modify_speaker_info.sh: copied data from data/train to exp/chain/ivectors_train/train_max2, number of speakers changed from 49989 to 49989
utils/validate_data_dir.sh: Successfully validated data-directory exp/chain/ivectors_train/train_max2
steps/online/nnet2/extract_ivectors_online.sh --cmd run.pl --nj 10 exp/chain/ivectors_train/train_max2 exp/chain/extractor exp/chain/ivectors_train
steps/online/nnet2/extract_ivectors_online.sh: extracting iVectors
steps/online/nnet2/extract_ivectors_online.sh: combining iVectors across jobs
steps/online/nnet2/extract_ivectors_online.sh: done extracting (online) iVectors to exp/chain/ivectors_train using the extractor in exp/chain/extractor.
local/chain/run_tdnn.sh: creating lang directory data/lang_chain with chain-type topology
steps/align_fmllr_lats.sh --nj 20 --cmd run.pl data/train data/lang exp/tri3 exp/chain/tri3_train_lats
steps/align_fmllr_lats.sh: feature type is lda
steps/align_fmllr_lats.sh: compiling training graphs
steps/align_fmllr_lats.sh: aligning data in data/train using exp/tri3/final.mdl and speaker-independent features.
steps/align_fmllr_lats.sh: computing fMLLR transforms
steps/align_fmllr_lats.sh: generating lattices containing alternate pronunciations.
steps/align_fmllr_lats.sh: done generating lattices from training transcripts.
4412 warnings in exp/chain/tri3_train_lats/log/generate_lattices.*.log
37396 warnings in exp/chain/tri3_train_lats/log/fmllr.*.log
18498 warnings in exp/chain/tri3_train_lats/log/align_pass1.*.log
steps/nnet3/chain/build_tree.sh --frame-subsampling-factor 3 --context-opts --context-width=2 --central-position=1 --cmd run.pl 2500 data/train data/lang_chain exp/tri3_ali exp/chain/tree
steps/nnet3/chain/build_tree.sh: feature type is lda
steps/nnet3/chain/build_tree.sh: Initializing monophone model (for alignment conversion, in case topology changed)
steps/nnet3/chain/build_tree.sh: Accumulating tree stats
steps/nnet3/chain/build_tree.sh: Getting questions for tree clustering.
steps/nnet3/chain/build_tree.sh: Building the tree
steps/nnet3/chain/build_tree.sh: Initializing the model
steps/nnet3/chain/build_tree.sh: Converting alignments from exp/tri3_ali to use current tree
steps/nnet3/chain/build_tree.sh: Done building tree
local/chain/run_tdnn.sh: creating neural net configs using the xconfig parser
tree-info exp/chain/tree/tree
steps/nnet3/xconfig_to_configs.py --xconfig-file exp/chain/tdnn/configs/network.xconfig --config-dir exp/chain/tdnn/configs/
nnet3-init exp/chain/tdnn/configs//ref.config exp/chain/tdnn/configs//ref.raw
LOG (nnet3-init[5.5.1012~1-dd107]:main():nnet3-init.cc:80) Initialized raw neural net and wrote it to exp/chain/tdnn/configs//ref.raw
nnet3-info exp/chain/tdnn/configs//ref.raw
nnet3-init exp/chain/tdnn/configs//ref.config exp/chain/tdnn/configs//ref.raw
LOG (nnet3-init[5.5.1012~1-dd107]:main():nnet3-init.cc:80) Initialized raw neural net and wrote it to exp/chain/tdnn/configs//ref.raw
nnet3-info exp/chain/tdnn/configs//ref.raw
/home/user/Documents/ML_Projects/vosk-api/kaldi/egs/wsj/training/steps/libs/common.py:127: SyntaxWarning: "is not" with a literal. Did you mean "!="?
if p.returncode is not 0:
/home/user/Documents/ML_Projects/vosk-api/kaldi/egs/wsj/training/steps/libs/common.py:147: SyntaxWarning: "is not" with a literal. Did you mean "!="?
if p.returncode is not 0:
/home/user/Documents/ML_Projects/vosk-api/kaldi/egs/wsj/training/steps/libs/common.py:203: SyntaxWarning: "is not" with a literal. Did you mean "!="?
if popen_object.returncode is not 0:
2022-05-12 21:43:32,423 [steps/nnet3/chain/train.py:35 - <module> - INFO ] Starting chain model trainer (train.py)
steps/nnet3/chain/train.py --stage -10 --cmd run.pl --feat.online-ivector-dir exp/chain/ivectors_train --feat.cmvn-opts --norm-means=false --norm-vars=false --chain.xent-regularize 0.1 --chain.leaky-hmm-coefficient 0.1 --chain.l2-regularize 0.0 --chain.apply-deriv-weights false --chain.lm-opts=--num-extra-lm-states=2000 --egs.cmd run.pl --egs.dir --egs.stage -10 --egs.opts --frames-overlap-per-eg 0 --constrained false --egs.chunk-width 140,100,160 --trainer.dropout-schedule 0,[email protected],[email protected],0 --trainer.add-option=--optimization.memory-compression-level=2 --trainer.num-chunk-per-minibatch 64 --trainer.frames-per-iter 2500000 --trainer.num-epochs 20 --trainer.optimization.num-jobs-initial 1 --trainer.optimization.num-jobs-final 1 --trainer.optimization.initial-effective-lrate 0.001 --trainer.optimization.final-effective-lrate 0.0001 --trainer.max-param-change 2.0 --cleanup.remove-egs true --feat-dir data/train --tree-dir exp/chain/tree --lat-dir exp/chain/tri3_train_lats --dir exp/chain/tdnn
['steps/nnet3/chain/train.py', '--stage', '-10', '--cmd', 'run.pl', '--feat.online-ivector-dir', 'exp/chain/ivectors_train', '--feat.cmvn-opts', '--norm-means=false --norm-vars=false', '--chain.xent-regularize', '0.1', '--chain.leaky-hmm-coefficient', '0.1', '--chain.l2-regularize', '0.0', '--chain.apply-deriv-weights', 'false', '--chain.lm-opts=--num-extra-lm-states=2000', '--egs.cmd', 'run.pl', '--egs.dir', '', '--egs.stage', '-10', '--egs.opts', '--frames-overlap-per-eg 0 --constrained false', '--egs.chunk-width', '140,100,160', '--trainer.dropout-schedule', '0,[email protected],[email protected],0', '--trainer.add-option=--optimization.memory-compression-level=2', '--trainer.num-chunk-per-minibatch', '64', '--trainer.frames-per-iter', '2500000', '--trainer.num-epochs', '20', '--trainer.optimization.num-jobs-initial', '1', '--trainer.optimization.num-jobs-final', '1', '--trainer.optimization.initial-effective-lrate', '0.001', '--trainer.optimization.final-effective-lrate', '0.0001', '--trainer.max-param-change', '2.0', '--cleanup.remove-egs', 'true', '--feat-dir', 'data/train', '--tree-dir', 'exp/chain/tree', '--lat-dir', 'exp/chain/tri3_train_lats', '--dir', 'exp/chain/tdnn']
2022-05-12 21:43:32,428 [steps/nnet3/chain/train.py:284 - train - INFO ] Arguments for the experiment
{'alignment_subsampling_factor': 3,
'apply_deriv_weights': False,
'backstitch_training_interval': 1,
'backstitch_training_scale': 0.0,
'chain_opts': '',
'chunk_left_context': 0,
'chunk_left_context_initial': -1,
'chunk_right_context': 0,
'chunk_right_context_final': -1,
'chunk_width': '140,100,160',
'cleanup': True,
'cmvn_opts': '--norm-means=false --norm-vars=false',
'combine_sum_to_one_penalty': 0.0,
'command': 'run.pl',
'compute_per_dim_accuracy': False,
'deriv_truncate_margin': None,
'dir': 'exp/chain/tdnn',
'do_final_combination': True,
'dropout_schedule': '0,[email protected],[email protected],0',
'egs_command': 'run.pl',
'egs_dir': None,
'egs_nj': 0,
'egs_opts': '--frames-overlap-per-eg 0 --constrained false',
'egs_stage': -10,
'email': None,
'exit_stage': None,
'feat_dir': 'data/train',
'final_effective_lrate': 0.0001,
'frame_subsampling_factor': 3,
'frames_per_iter': 2500000,
'initial_effective_lrate': 0.001,
'input_model': None,
'l2_regularize': 0.0,
'lat_dir': 'exp/chain/tri3_train_lats',
'leaky_hmm_coefficient': 0.1,
'left_deriv_truncate': None,
'left_tolerance': 5,
'lm_opts': '--num-extra-lm-states=2000',
'max_lda_jobs': 10,
'max_models_combine': 20,
'max_objective_evaluations': 30,
'max_param_change': 2.0,
'momentum': 0.0,
'num_chunk_per_minibatch': '64',
'num_epochs': 20.0,
'num_jobs_final': 1,
'num_jobs_initial': 1,
'num_jobs_step': 1,
'online_ivector_dir': 'exp/chain/ivectors_train',
'preserve_model_interval': 100,
'presoftmax_prior_scale_power': -0.25,
'proportional_shrink': 0.0,
'rand_prune': 4.0,
'remove_egs': True,
'reporting_interval': 0.1,
'right_tolerance': 5,
'samples_per_iter': 400000,
'shrink_saturation_threshold': 0.4,
'shrink_value': 1.0,
'shuffle_buffer_size': 5000,
'srand': 0,
'stage': -10,
'train_opts': ['--optimization.memory-compression-level=2'],
'tree_dir': 'exp/chain/tree',
'use_gpu': 'yes',
'xent_regularize': 0.1}
2022-05-12 21:43:32,456 [steps/nnet3/chain/train.py:341 - train - INFO ] Creating phone language-model
2022-05-12 21:43:34,363 [steps/nnet3/chain/train.py:346 - train - INFO ] Creating denominator FST
copy-transition-model exp/chain/tree/final.mdl exp/chain/tdnn/0.trans_mdl
LOG (copy-transition-model[5.5.1012~1-dd107]:main():copy-transition-model.cc:62) Copied transition model.
2022-05-12 21:43:35,035 [steps/nnet3/chain/train.py:382 - train - INFO ] Generating egs
steps/nnet3/chain/get_egs.sh --frames-overlap-per-eg 0 --constrained false --cmd run.pl --cmvn-opts --norm-means=false --norm-vars=false --online-ivector-dir exp/chain/ivectors_train --left-context 27 --right-context 27 --left-context-initial -1 --right-context-final -1 --left-tolerance 5 --right-tolerance 5 --frame-subsampling-factor 3 --alignment-subsampling-factor 3 --stage -10 --frames-per-iter 2500000 --frames-per-eg 140,100,160 --srand 0 data/train exp/chain/tdnn exp/chain/tri3_train_lats exp/chain/tdnn/egs
steps/nnet3/chain/get_egs.sh: Holding out 300 utterances in validation set and 300 in training diagnostic set, out of total 49989.
steps/nnet3/chain/get_egs.sh: creating egs. To ensure they are not deleted later you can do: touch exp/chain/tdnn/egs/.nodelete
steps/nnet3/chain/get_egs.sh: feature type is raw, with 'apply-cmvn'
tree-info exp/chain/tdnn/tree
feat-to-dim scp:exp/chain/ivectors_train/ivector_online.scp -
steps/nnet3/chain/get_egs.sh: working out number of frames of training data
steps/nnet3/chain/get_egs.sh: working out feature dim
steps/nnet3/chain/get_egs.sh: creating 16 archives, each with 14985 egs, with
steps/nnet3/chain/get_egs.sh: 140,100,160 labels per example, and (left,right) context = (27,27)
steps/nnet3/chain/get_egs.sh: Getting validation and training subset examples in background.
steps/nnet3/chain/get_egs.sh: Generating training examples on disk
steps/nnet3/chain/get_egs.sh: Getting subsets of validation examples for diagnostics and combination.
steps/nnet3/chain/get_egs.sh: recombining and shuffling order of archives on disk
steps/nnet3/chain/get_egs.sh: Removing temporary archives, alignments and lattices
steps/nnet3/chain/get_egs.sh: Finished preparing training examples
2022-05-12 21:44:48,469 [steps/nnet3/chain/train.py:431 - train - INFO ] Copying the properties from exp/chain/tdnn/egs to exp/chain/tdnn
2022-05-12 21:44:48,469 [steps/nnet3/chain/train.py:454 - train - INFO ] Preparing the initial acoustic model.
2022-05-12 21:44:48,739 [steps/nnet3/chain/train.py:487 - train - INFO ] Training will run for 20.0 epochs = 960 iterations
2022-05-12 21:44:48,739 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 0/959 Jobs: 1 Epoch: 0.00/20.0 (0.0% complete) lr: 0.001000
2022-05-12 21:45:14,288 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 1/959 Jobs: 1 Epoch: 0.02/20.0 (0.1% complete) lr: 0.000998
2022-05-12 21:45:34,829 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 2/959 Jobs: 1 Epoch: 0.04/20.0 (0.2% complete) lr: 0.000995
2022-05-12 21:45:55,693 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 3/959 Jobs: 1 Epoch: 0.06/20.0 (0.3% complete) lr: 0.000993
2022-05-12 21:46:16,417 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 4/959 Jobs: 1 Epoch: 0.08/20.0 (0.4% complete) lr: 0.000990
2022-05-12 21:46:37,271 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 5/959 Jobs: 1 Epoch: 0.10/20.0 (0.5% complete) lr: 0.000988
2022-05-12 21:46:58,264 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 6/959 Jobs: 1 Epoch: 0.12/20.0 (0.6% complete) lr: 0.000986
2022-05-12 21:47:18,914 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 7/959 Jobs: 1 Epoch: 0.15/20.0 (0.7% complete) lr: 0.000983
2022-05-12 21:47:40,045 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 8/959 Jobs: 1 Epoch: 0.17/20.0 (0.8% complete) lr: 0.000981
2022-05-12 21:48:00,772 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 9/959 Jobs: 1 Epoch: 0.19/20.0 (0.9% complete) lr: 0.000979
2022-05-12 21:48:21,401 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 10/959 Jobs: 1 Epoch: 0.21/20.0 (1.0% complete) lr: 0.000976
2022-05-12 21:48:42,254 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 11/959 Jobs: 1 Epoch: 0.23/20.0 (1.1% complete) lr: 0.000974
2022-05-12 21:49:02,967 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 12/959 Jobs: 1 Epoch: 0.25/20.0 (1.2% complete) lr: 0.000972
2022-05-12 21:49:23,566 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 13/959 Jobs: 1 Epoch: 0.27/20.0 (1.4% complete) lr: 0.000969
2022-05-12 21:49:44,214 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 14/959 Jobs: 1 Epoch: 0.29/20.0 (1.5% complete) lr: 0.000967
2022-05-12 21:50:05,064 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 15/959 Jobs: 1 Epoch: 0.31/20.0 (1.6% complete) lr: 0.000965
2022-05-12 21:50:25,853 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 16/959 Jobs: 1 Epoch: 0.33/20.0 (1.7% complete) lr: 0.000962
2022-05-12 21:50:46,246 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 17/959 Jobs: 1 Epoch: 0.35/20.0 (1.8% complete) lr: 0.000960
2022-05-12 21:51:07,165 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 18/959 Jobs: 1 Epoch: 0.38/20.0 (1.9% complete) lr: 0.000958
2022-05-12 21:51:28,136 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 19/959 Jobs: 1 Epoch: 0.40/20.0 (2.0% complete) lr: 0.000955
2022-05-12 21:51:48,700 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 20/959 Jobs: 1 Epoch: 0.42/20.0 (2.1% complete) lr: 0.000953
2022-05-12 21:52:15,183 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 21/959 Jobs: 1 Epoch: 0.44/20.0 (2.2% complete) lr: 0.000951
2022-05-12 21:52:36,342 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 22/959 Jobs: 1 Epoch: 0.46/20.0 (2.3% complete) lr: 0.000949
2022-05-12 21:52:57,016 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 23/959 Jobs: 1 Epoch: 0.48/20.0 (2.4% complete) lr: 0.000946
2022-05-12 21:53:17,775 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 24/959 Jobs: 1 Epoch: 0.50/20.0 (2.5% complete) lr: 0.000944
2022-05-12 21:53:38,530 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 25/959 Jobs: 1 Epoch: 0.52/20.0 (2.6% complete) lr: 0.000942
2022-05-12 21:53:59,345 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 26/959 Jobs: 1 Epoch: 0.54/20.0 (2.7% complete) lr: 0.000940
2022-05-12 21:54:19,954 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 27/959 Jobs: 1 Epoch: 0.56/20.0 (2.8% complete) lr: 0.000937
2022-05-12 21:54:40,398 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 28/959 Jobs: 1 Epoch: 0.58/20.0 (2.9% complete) lr: 0.000935
2022-05-12 21:55:01,123 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 29/959 Jobs: 1 Epoch: 0.60/20.0 (3.0% complete) lr: 0.000933
2022-05-12 21:55:21,882 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 30/959 Jobs: 1 Epoch: 0.62/20.0 (3.1% complete) lr: 0.000931
2022-05-12 21:55:42,407 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 31/959 Jobs: 1 Epoch: 0.65/20.0 (3.2% complete) lr: 0.000928
2022-05-12 21:56:03,175 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 32/959 Jobs: 1 Epoch: 0.67/20.0 (3.3% complete) lr: 0.000926
2022-05-12 21:56:23,764 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 33/959 Jobs: 1 Epoch: 0.69/20.0 (3.4% complete) lr: 0.000924
2022-05-12 21:56:44,308 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 34/959 Jobs: 1 Epoch: 0.71/20.0 (3.5% complete) lr: 0.000922
2022-05-12 21:57:04,985 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 35/959 Jobs: 1 Epoch: 0.73/20.0 (3.6% complete) lr: 0.000919
2022-05-12 21:57:25,656 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 36/959 Jobs: 1 Epoch: 0.75/20.0 (3.8% complete) lr: 0.000917
2022-05-12 21:57:46,615 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 37/959 Jobs: 1 Epoch: 0.77/20.0 (3.9% complete) lr: 0.000915
2022-05-12 21:58:07,497 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 38/959 Jobs: 1 Epoch: 0.79/20.0 (4.0% complete) lr: 0.000913
2022-05-12 21:58:28,079 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 39/959 Jobs: 1 Epoch: 0.81/20.0 (4.1% complete) lr: 0.000911
2022-05-12 21:58:49,089 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 40/959 Jobs: 1 Epoch: 0.83/20.0 (4.2% complete) lr: 0.000909
2022-05-12 21:59:15,382 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 41/959 Jobs: 1 Epoch: 0.85/20.0 (4.3% complete) lr: 0.000906
2022-05-12 21:59:35,995 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 42/959 Jobs: 1 Epoch: 0.88/20.0 (4.4% complete) lr: 0.000904
2022-05-12 21:59:56,837 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 43/959 Jobs: 1 Epoch: 0.90/20.0 (4.5% complete) lr: 0.000902
2022-05-12 22:00:17,523 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 44/959 Jobs: 1 Epoch: 0.92/20.0 (4.6% complete) lr: 0.000900
2022-05-12 22:00:37,961 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 45/959 Jobs: 1 Epoch: 0.94/20.0 (4.7% complete) lr: 0.000898
2022-05-12 22:00:58,855 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 46/959 Jobs: 1 Epoch: 0.96/20.0 (4.8% complete) lr: 0.000896
2022-05-12 22:01:19,600 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 47/959 Jobs: 1 Epoch: 0.98/20.0 (4.9% complete) lr: 0.000893
2022-05-12 22:01:40,237 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 48/959 Jobs: 1 Epoch: 1.00/20.0 (5.0% complete) lr: 0.000891
2022-05-12 22:02:00,557 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 49/959 Jobs: 1 Epoch: 1.02/20.0 (5.1% complete) lr: 0.000889
2022-05-12 22:02:21,159 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 50/959 Jobs: 1 Epoch: 1.04/20.0 (5.2% complete) lr: 0.000887
2022-05-12 22:02:42,075 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 51/959 Jobs: 1 Epoch: 1.06/20.0 (5.3% complete) lr: 0.000885
2022-05-12 22:03:02,544 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 52/959 Jobs: 1 Epoch: 1.08/20.0 (5.4% complete) lr: 0.000883
2022-05-12 22:03:23,218 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 53/959 Jobs: 1 Epoch: 1.10/20.0 (5.5% complete) lr: 0.000881
2022-05-12 22:03:44,316 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 54/959 Jobs: 1 Epoch: 1.12/20.0 (5.6% complete) lr: 0.000879
2022-05-12 22:04:04,880 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 55/959 Jobs: 1 Epoch: 1.15/20.0 (5.7% complete) lr: 0.000876
2022-05-12 22:04:25,628 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 56/959 Jobs: 1 Epoch: 1.17/20.0 (5.8% complete) lr: 0.000874
2022-05-12 22:04:46,324 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 57/959 Jobs: 1 Epoch: 1.19/20.0 (5.9% complete) lr: 0.000872
2022-05-12 22:05:07,063 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 58/959 Jobs: 1 Epoch: 1.21/20.0 (6.0% complete) lr: 0.000870
2022-05-12 22:05:27,660 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 59/959 Jobs: 1 Epoch: 1.23/20.0 (6.1% complete) lr: 0.000868
2022-05-12 22:05:48,131 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 60/959 Jobs: 1 Epoch: 1.25/20.0 (6.2% complete) lr: 0.000866
2022-05-12 22:06:14,326 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 61/959 Jobs: 1 Epoch: 1.27/20.0 (6.4% complete) lr: 0.000864
2022-05-12 22:06:35,122 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 62/959 Jobs: 1 Epoch: 1.29/20.0 (6.5% complete) lr: 0.000862
2022-05-12 22:06:55,651 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 63/959 Jobs: 1 Epoch: 1.31/20.0 (6.6% complete) lr: 0.000860
2022-05-12 22:07:16,332 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 64/959 Jobs: 1 Epoch: 1.33/20.0 (6.7% complete) lr: 0.000858
2022-05-12 22:07:36,802 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 65/959 Jobs: 1 Epoch: 1.35/20.0 (6.8% complete) lr: 0.000856
2022-05-12 22:07:57,249 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 66/959 Jobs: 1 Epoch: 1.38/20.0 (6.9% complete) lr: 0.000854
2022-05-12 22:08:17,997 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 67/959 Jobs: 1 Epoch: 1.40/20.0 (7.0% complete) lr: 0.000852
2022-05-12 22:08:38,643 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 68/959 Jobs: 1 Epoch: 1.42/20.0 (7.1% complete) lr: 0.000850
2022-05-12 22:08:59,452 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 69/959 Jobs: 1 Epoch: 1.44/20.0 (7.2% complete) lr: 0.000847
2022-05-12 22:09:20,254 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 70/959 Jobs: 1 Epoch: 1.46/20.0 (7.3% complete) lr: 0.000845
2022-05-12 22:09:40,952 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 71/959 Jobs: 1 Epoch: 1.48/20.0 (7.4% complete) lr: 0.000843
2022-05-12 22:10:01,898 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 72/959 Jobs: 1 Epoch: 1.50/20.0 (7.5% complete) lr: 0.000841
2022-05-12 22:10:22,491 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 73/959 Jobs: 1 Epoch: 1.52/20.0 (7.6% complete) lr: 0.000839
2022-05-12 22:10:43,057 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 74/959 Jobs: 1 Epoch: 1.54/20.0 (7.7% complete) lr: 0.000837
2022-05-12 22:11:03,800 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 75/959 Jobs: 1 Epoch: 1.56/20.0 (7.8% complete) lr: 0.000835
2022-05-12 22:11:24,409 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 76/959 Jobs: 1 Epoch: 1.58/20.0 (7.9% complete) lr: 0.000833
2022-05-12 22:11:44,842 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 77/959 Jobs: 1 Epoch: 1.60/20.0 (8.0% complete) lr: 0.000831
2022-05-12 22:12:05,681 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 78/959 Jobs: 1 Epoch: 1.62/20.0 (8.1% complete) lr: 0.000829
2022-05-12 22:12:26,355 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 79/959 Jobs: 1 Epoch: 1.65/20.0 (8.2% complete) lr: 0.000827
2022-05-12 22:12:47,024 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 80/959 Jobs: 1 Epoch: 1.67/20.0 (8.3% complete) lr: 0.000825
2022-05-12 22:13:12,798 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 81/959 Jobs: 1 Epoch: 1.69/20.0 (8.4% complete) lr: 0.000823
2022-05-12 22:13:33,400 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 82/959 Jobs: 1 Epoch: 1.71/20.0 (8.5% complete) lr: 0.000821
2022-05-12 22:13:54,179 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 83/959 Jobs: 1 Epoch: 1.73/20.0 (8.6% complete) lr: 0.000819
2022-05-12 22:14:14,530 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 84/959 Jobs: 1 Epoch: 1.75/20.0 (8.8% complete) lr: 0.000818
2022-05-12 22:14:35,526 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 85/959 Jobs: 1 Epoch: 1.77/20.0 (8.9% complete) lr: 0.000816
2022-05-12 22:14:56,695 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 86/959 Jobs: 1 Epoch: 1.79/20.0 (9.0% complete) lr: 0.000814
2022-05-12 22:15:17,387 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 87/959 Jobs: 1 Epoch: 1.81/20.0 (9.1% complete) lr: 0.000812
2022-05-12 22:15:38,198 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 88/959 Jobs: 1 Epoch: 1.83/20.0 (9.2% complete) lr: 0.000810
2022-05-12 22:15:58,896 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 89/959 Jobs: 1 Epoch: 1.85/20.0 (9.3% complete) lr: 0.000808
2022-05-12 22:16:19,594 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 90/959 Jobs: 1 Epoch: 1.88/20.0 (9.4% complete) lr: 0.000806
2022-05-12 22:16:40,171 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 91/959 Jobs: 1 Epoch: 1.90/20.0 (9.5% complete) lr: 0.000804
2022-05-12 22:17:00,646 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 92/959 Jobs: 1 Epoch: 1.92/20.0 (9.6% complete) lr: 0.000802
2022-05-12 22:17:21,306 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 93/959 Jobs: 1 Epoch: 1.94/20.0 (9.7% complete) lr: 0.000800
2022-05-12 22:17:42,093 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 94/959 Jobs: 1 Epoch: 1.96/20.0 (9.8% complete) lr: 0.000798
2022-05-12 22:18:02,595 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 95/959 Jobs: 1 Epoch: 1.98/20.0 (9.9% complete) lr: 0.000796
2022-05-12 22:18:23,277 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 96/959 Jobs: 1 Epoch: 2.00/20.0 (10.0% complete) lr: 0.000794
2022-05-12 22:18:43,815 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 97/959 Jobs: 1 Epoch: 2.02/20.0 (10.1% complete) lr: 0.000792
2022-05-12 22:19:04,340 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 98/959 Jobs: 1 Epoch: 2.04/20.0 (10.2% complete) lr: 0.000791
2022-05-12 22:19:25,054 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 99/959 Jobs: 1 Epoch: 2.06/20.0 (10.3% complete) lr: 0.000789
2022-05-12 22:19:45,718 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 100/959 Jobs: 1 Epoch: 2.08/20.0 (10.4% complete) lr: 0.000787
2022-05-12 22:20:12,210 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 101/959 Jobs: 1 Epoch: 2.10/20.0 (10.5% complete) lr: 0.000785
2022-05-12 22:20:33,074 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 102/959 Jobs: 1 Epoch: 2.12/20.0 (10.6% complete) lr: 0.000783
2022-05-12 22:20:53,819 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 103/959 Jobs: 1 Epoch: 2.15/20.0 (10.7% complete) lr: 0.000781
2022-05-12 22:21:14,806 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 104/959 Jobs: 1 Epoch: 2.17/20.0 (10.8% complete) lr: 0.000779
2022-05-12 22:21:35,385 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 105/959 Jobs: 1 Epoch: 2.19/20.0 (10.9% complete) lr: 0.000777
2022-05-12 22:21:55,918 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 106/959 Jobs: 1 Epoch: 2.21/20.0 (11.0% complete) lr: 0.000776
2022-05-12 22:22:16,675 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 107/959 Jobs: 1 Epoch: 2.23/20.0 (11.1% complete) lr: 0.000774
2022-05-12 22:22:37,244 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 108/959 Jobs: 1 Epoch: 2.25/20.0 (11.2% complete) lr: 0.000772
2022-05-12 22:22:57,666 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 109/959 Jobs: 1 Epoch: 2.27/20.0 (11.4% complete) lr: 0.000770
2022-05-12 22:23:18,464 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 110/959 Jobs: 1 Epoch: 2.29/20.0 (11.5% complete) lr: 0.000768
2022-05-12 22:23:39,229 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 111/959 Jobs: 1 Epoch: 2.31/20.0 (11.6% complete) lr: 0.000766
2022-05-12 22:23:59,816 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 112/959 Jobs: 1 Epoch: 2.33/20.0 (11.7% complete) lr: 0.000764
2022-05-12 22:24:20,175 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 113/959 Jobs: 1 Epoch: 2.35/20.0 (11.8% complete) lr: 0.000763
2022-05-12 22:24:40,799 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 114/959 Jobs: 1 Epoch: 2.38/20.0 (11.9% complete) lr: 0.000761
2022-05-12 22:25:01,584 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 115/959 Jobs: 1 Epoch: 2.40/20.0 (12.0% complete) lr: 0.000759
2022-05-12 22:25:22,011 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 116/959 Jobs: 1 Epoch: 2.42/20.0 (12.1% complete) lr: 0.000757
2022-05-12 22:25:42,774 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 117/959 Jobs: 1 Epoch: 2.44/20.0 (12.2% complete) lr: 0.000755
2022-05-13 03:16:25,873 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 938/959 Jobs: 1 Epoch: 19.54/20.0 (97.7% complete) lr: 0.000105
2022-05-13 03:16:46,805 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 939/959 Jobs: 1 Epoch: 19.56/20.0 (97.8% complete) lr: 0.000105
2022-05-13 03:17:07,547 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 940/959 Jobs: 1 Epoch: 19.58/20.0 (97.9% complete) lr: 0.000105
2022-05-13 03:17:34,074 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 941/959 Jobs: 1 Epoch: 19.60/20.0 (98.0% complete) lr: 0.000105
2022-05-13 03:17:55,188 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 942/959 Jobs: 1 Epoch: 19.62/20.0 (98.1% complete) lr: 0.000104
2022-05-13 03:18:16,078 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 943/959 Jobs: 1 Epoch: 19.65/20.0 (98.2% complete) lr: 0.000104
2022-05-13 03:18:36,844 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 944/959 Jobs: 1 Epoch: 19.67/20.0 (98.3% complete) lr: 0.000104
2022-05-13 03:18:57,689 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 945/959 Jobs: 1 Epoch: 19.69/20.0 (98.4% complete) lr: 0.000104
2022-05-13 03:19:18,545 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 946/959 Jobs: 1 Epoch: 19.71/20.0 (98.5% complete) lr: 0.000103
2022-05-13 03:19:39,511 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 947/959 Jobs: 1 Epoch: 19.73/20.0 (98.6% complete) lr: 0.000103
2022-05-13 03:20:00,370 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 948/959 Jobs: 1 Epoch: 19.75/20.0 (98.8% complete) lr: 0.000103
2022-05-13 03:20:21,540 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 949/959 Jobs: 1 Epoch: 19.77/20.0 (98.9% complete) lr: 0.000103
2022-05-13 03:20:42,841 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 950/959 Jobs: 1 Epoch: 19.79/20.0 (99.0% complete) lr: 0.000102
2022-05-13 03:21:03,773 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 951/959 Jobs: 1 Epoch: 19.81/20.0 (99.1% complete) lr: 0.000102
2022-05-13 03:21:25,128 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 952/959 Jobs: 1 Epoch: 19.83/20.0 (99.2% complete) lr: 0.000102
2022-05-13 03:21:46,070 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 953/959 Jobs: 1 Epoch: 19.85/20.0 (99.3% complete) lr: 0.000102
2022-05-13 03:22:06,945 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 954/959 Jobs: 1 Epoch: 19.88/20.0 (99.4% complete) lr: 0.000101
2022-05-13 03:22:27,827 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 955/959 Jobs: 1 Epoch: 19.90/20.0 (99.5% complete) lr: 0.000101
2022-05-13 03:22:48,833 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 956/959 Jobs: 1 Epoch: 19.92/20.0 (99.6% complete) lr: 0.000101
2022-05-13 03:23:09,761 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 957/959 Jobs: 1 Epoch: 19.94/20.0 (99.7% complete) lr: 0.000101
2022-05-13 03:23:30,733 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 958/959 Jobs: 1 Epoch: 19.96/20.0 (99.8% complete) lr: 0.000100
2022-05-13 03:23:51,775 [steps/nnet3/chain/train.py:529 - train - INFO ] Iter: 959/959 Jobs: 1 Epoch: 19.98/20.0 (99.9% complete) lr: 0.000100
2022-05-13 03:24:12,736 [steps/nnet3/chain/train.py:592 - train - INFO ] Doing final combination to produce final.mdl
2022-05-13 03:24:12,736 [/home/user/Documents/ML_Projects/vosk-api/kaldi/egs/wsj/training/steps/libs/nnet3/train/chain_objf/acoustic_model.py:571 - combine_models - INFO ] Combining {936, 937, 938, 939, 940, 941, 942, 943, 944, 945, 946, 947, 948, 949, 950, 951, 952, 953, 954, 955, 956, 957, 958, 959, 960} models.
2022-05-13 03:24:36,944 [steps/nnet3/chain/train.py:620 - train - INFO ] Cleaning up the experiment directory exp/chain/tdnn
steps/nnet2/remove_egs.sh: Finished deleting examples in exp/chain/tdnn/egs
exp/chain/tdnn: num-iters=960 nj=1..1 num-params=3.5M dim=40+40->2176 combine=-0.127->-0.119 (over 13) xent:train/valid[638,959]=(-2.30,-2.09/-2.34,-2.15) logprob:train/valid[638,959]=(-0.147,-0.126/-0.172,-0.152)
Converting 'data/local/lm/lm_tgsmall.arpa.gz' to FST
arpa2fst --disambig-symbol=#0 --read-symbol-table=data/lang_test/words.txt - data/lang_test/G.fst
LOG (arpa2fst[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:94) Reading \data\ section.
LOG (arpa2fst[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \1-grams: section.
LOG (arpa2fst[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \2-grams: section.
LOG (arpa2fst[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \3-grams: section.
LOG (arpa2fst[5.5.1012~1-dd107]:RemoveRedundantStates():arpa-lm-compiler.cc:359) Reduced num-states from 1183418 to 249533
fstisstochastic data/lang_test/G.fst
9.90926e-06 -0.29805
Succeeded in formatting LM: 'data/local/lm/lm_tgsmall.arpa.gz'
tree-info exp/chain/tdnn/tree
tree-info exp/chain/tdnn/tree
fstpushspecial
fstminimizeencoded
fstdeterminizestar --use-log=true
fsttablecompose data/lang_test/L_disambig.fst data/lang_test/G.fst
fstisstochastic data/lang_test/tmp/LG.fst
-0.0577829 -0.0586653
[info]: LG not stochastic.
fstcomposecontext --context-size=2 --central-position=1 --read-disambig-syms=data/lang_test/phones/disambig.int --write-disambig-syms=data/lang_test/tmp/disambig_ilabels_2_1.int data/lang_test/tmp/ilabels_2_1.1396775 data/lang_test/tmp/LG.fst
fstisstochastic data/lang_test/tmp/CLG_2_1.fst
-0.0577829 -0.0586653
[info]: CLG not stochastic.
make-h-transducer --disambig-syms-out=exp/chain/tdnn/graph/disambig_tid.int --transition-scale=1.0 data/lang_test/tmp/ilabels_2_1 exp/chain/tdnn/tree exp/chain/tdnn/final.mdl
fsttablecompose exp/chain/tdnn/graph/Ha.fst data/lang_test/tmp/CLG_2_1.fst
fstminimizeencoded
fstdeterminizestar --use-log=true
fstrmsymbols exp/chain/tdnn/graph/disambig_tid.int
fstrmepslocal
fstisstochastic exp/chain/tdnn/graph/HCLGa.fst
-0.0191171 -0.213417
HCLGa is not stochastic
add-self-loops --self-loop-scale=1.0 --reorder=true exp/chain/tdnn/final.mdl exp/chain/tdnn/graph/HCLGa.fst
fstisstochastic exp/chain/tdnn/graph/HCLG.fst
1.90465e-09 -0.146711
[info]: final HCLG is not stochastic.
arpa-to-const-arpa --bos-symbol=200004 --eos-symbol=200005 --unk-symbol=2 'gunzip -c data/local/lm/lm_tgmed.arpa.gz | utils/map_arpa_lm.pl data/lang_test_rescore/words.txt|' data/lang_test_rescore/G.carpa
LOG (arpa-to-const-arpa[5.5.1012~1-dd107]:BuildConstArpaLm():const-arpa-lm.cc:1078) Reading gunzip -c data/local/lm/lm_tgmed.arpa.gz | utils/map_arpa_lm.pl data/lang_test_rescore/words.txt|
utils/map_arpa_lm.pl: Processing "\data\"
utils/map_arpa_lm.pl: Processing "\1-grams:\"
LOG (arpa-to-const-arpa[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:94) Reading \data\ section.
LOG (arpa-to-const-arpa[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \1-grams: section.
utils/map_arpa_lm.pl: Processing "\2-grams:\"
LOG (arpa-to-const-arpa[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \2-grams: section.
utils/map_arpa_lm.pl: Processing "\3-grams:\"
LOG (arpa-to-const-arpa[5.5.1012~1-dd107]:Read():arpa-file-parser.cc:149) Reading \3-grams: section.
steps/make_mfcc.sh --cmd run.pl --nj 10 data/test exp/make_mfcc/test
steps/make_mfcc.sh: moving data/test/feats.scp to data/test/.backup
utils/validate_data_dir.sh: Successfully validated data-directory data/test
steps/make_mfcc.sh: [info]: no segments file exists: assuming wav.scp indexed by utterance.
steps/make_mfcc.sh: Succeeded creating MFCC features for test
steps/compute_cmvn_stats.sh data/test exp/make_mfcc/test
Succeeded creating CMVN stats for test
steps/online/nnet2/extract_ivectors_online.sh --nj 10 data/test exp/chain/extractor exp/chain/ivectors_test
steps/online/nnet2/extract_ivectors_online.sh: extracting iVectors
steps/online/nnet2/extract_ivectors_online.sh: combining iVectors across jobs
steps/online/nnet2/extract_ivectors_online.sh: done extracting (online) iVectors to exp/chain/ivectors_test using the extractor in exp/chain/extractor.
steps/nnet3/decode.sh --cmd run.pl --num-threads 10 --nj 1 --beam 13.0 --max-active 7000 --lattice-beam 4.0 --online-ivector-dir exp/chain/ivectors_test --acwt 1.0 --post-decode-acwt 10.0 exp/chain/tdnn/graph data/test exp/chain/tdnn/decode_test
steps/nnet3/decode.sh: feature type is raw
steps/diagnostic/analyze_lats.sh --cmd run.pl --iter final exp/chain/tdnn/graph exp/chain/tdnn/decode_test
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 19.5202944648% of the time at utterance begin. This may not be optimal.
analyze_phone_length_stats.py: WARNING: optional-silence SIL is seen only 75.1856225109% of the time at utterance end. This may not be optimal.
steps/diagnostic/analyze_lats.sh: see stats in exp/chain/tdnn/decode_test/log/analyze_alignments.log
Overall, lattice depth (10,50,90-percentile)=(1,5,27) and mean=11.0
steps/diagnostic/analyze_lats.sh: see stats in exp/chain/tdnn/decode_test/log/analyze_lattice_depth_stats.log
score best paths
local/score.sh --cmd run.pl data/test exp/chain/tdnn/graph exp/chain/tdnn/decode_test
local/score.sh: scoring with word insertion penalty=0.0
score confidence and timing with sclite
Decoding done.
steps/lmrescore_const_arpa.sh data/lang_test data/lang_test_rescore data/test exp/chain/tdnn/decode_test exp/chain/tdnn/decode_test_rescore
local/score.sh --cmd run.pl data/test data/lang_test_rescore exp/chain/tdnn/decode_test_rescore
local/score.sh: scoring with word insertion penalty=0.0
grep: exp/tri3b/decode_tgsmall_test/wer_*: No such file or directory
%WER 36.23 [ 313658 / 865805, 57762 ins, 92680 del, 163216 sub ] exp/chain/tdnn/decode_test/wer_7_0.0
%WER 35.78 [ 309776 / 865805, 57306 ins, 92931 del, 159539 sub ] exp/chain/tdnn/decode_test_rescore/wer_7_0.0
Please tell me which files i need to create model, please tell me steps.
Submit a patch to put the model with your initial understanding, we will correct.
Please tell me which files i need to create model, please tell me steps.
Submit a patch to put the model with your initial understanding, we will correct.
Sorry sir, but i can't understand what you said,