Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
682 commits
Select commit Hold shift + click to select a range
e922333
[egs] New chime-5 recipe (#2893)
vimalmanohar May 12, 2019
a861e56
[scripts,egs] Made changes to the augmentation script to make it work…
phanisankar-nidadavolu May 13, 2019
cec8958
[egs] updated local/musan.sh to steps/data/make_musan.sh in speaker i…
phanisankar-nidadavolu May 13, 2019
d40222e
[src] Fix sample rounding errors in extract-segments (#3321)
May 14, 2019
35f96db
[src,scripts]Store frame_shift, utt2{dur,num_frames}, .conf with feat…
May 14, 2019
a2e7ba3
[build] Initial version of Docker images for (CPU and GPU versions) (…
mdoulaty May 15, 2019
91609c7
[scripts] fix typo/bug in make_musan.py (#3327)
May 15, 2019
95e81c0
[scripts] Fixed misnamed variable in data/make_musan.py (#3324)
phanisankar-nidadavolu May 15, 2019
c5aa3a9
[scripts] Trust frame_shift and utt2num_frames if found (#3313)
May 16, 2019
0ff318b
[scripts] typo fix in augmentation script (#3329)
May 16, 2019
62ebb44
[scripts] handle frame_shit and utt2num_frames in utils/ (#3323)
May 16, 2019
c8b93bc
[scripts] Extend combine_ali_dirs.sh to combine alignment lattices (#…
May 17, 2019
528e072
[src] Fix rare case when segment end rounding overshoots file end in …
alumae May 17, 2019
8397e05
[scripts] Change --modify-spk-id default to False; back-compatibility…
phanisankar-nidadavolu May 20, 2019
8b54ef8
[build] Add easier configure option in failure message of configure (…
danpovey May 20, 2019
ce8798b
[scripts,minor] Fix typo in comment (#3338)
Shujian2015 May 22, 2019
9e0a7f6
[src,egs] Add option for applying SVD on trained models (#3272)
saikiranvalluri May 23, 2019
0e5e07b
[src] Add interfaces to nnet-batch-compute that expects device input.…
luitjens May 23, 2019
52e7ecf
[build] Update GCC support check for CUDA toolkit 10.1 (#3345)
entn-at May 27, 2019
29f3c14
[egs] Fix to aishell1 v1 download script (#3344)
naxingyu May 27, 2019
a5dd6bd
[scripts] Support utf-8 files in some scripts (#3346)
vimalmanohar May 28, 2019
8c6cd31
[src] Fix potential underflow bug in MFCC, RE energy floor, thx: Zolt…
huangruizhe May 28, 2019
e643c73
[scripts]: add warning to nnet3/chain/train.py about ineffective opti…
bringtree May 28, 2019
8706f06
[scripts] Fix regarding UTF handling in cleanup script (#3352)
vimalmanohar May 29, 2019
800924d
[scripts] Change encoding to utf-8 in data augmentation scripts (#3360)
hhadian Jun 1, 2019
eedd9fa
[src] Add CUDA accelerated MFCC computation. (#3348)
luitjens Jun 3, 2019
0b443bd
[src] Optimizations for batch nnet3. The issue fixed here is that (#…
luitjens Jun 3, 2019
16097b4
[scripts,minor] Remove outdated comment (#3361)
Shujian2015 Jun 3, 2019
ced53e1
[egs] A kaldi recipe based on the corpus named "aidatatang_200zh". (#…
DatatangAI Jun 4, 2019
f8a4376
[src] nnet1: changing end-rule in 'nnet-train-multistream', (#3358)
KarelVesely84 Jun 4, 2019
9c734a5
[scripts] Fix how the empty (faulty?) segments are handled in data-cl…
jtrmal Jun 4, 2019
b276d70
[src] Fix to bug in ivector extraction causing assert failure, thx: s…
danpovey Jun 4, 2019
de4a3e3
[src] Fix to bug in ivector extraction causing assert failure, thx: s…
danpovey Jun 4, 2019
1a4aa52
[scripts] add script to compute dev PPL on kaldi-rnnlm (#3340)
hainan-xv Jun 4, 2019
1735003
[scripts,egs] Small fixes to diarization scripts (#3366)
HuangZiliAndy Jun 4, 2019
338cc58
[egs] Modify split_scp.pl usage to match its updated code (#3371)
danpovey Jun 5, 2019
254d636
[src] Fix non-cuda `make depend` build by putting compile guards arou…
luitjens Jun 6, 2019
3648df5
[build] Docker docs update and minor changes to the Docker files (#3…
mdoulaty Jun 6, 2019
0071003
[egs] Scripts for MATERIAL ASR (#2165)
mahsa7823 Jun 6, 2019
acff3f6
[src] Batch nnet3 optimizations. Batch some of the copies in and cop…
luitjens Jun 6, 2019
23ba982
[build] Widen cuda guard in cudafeat makefile. (#3379)
langep Jun 7, 2019
04cf43b
[scripts] nnet1: updating the scripts to support 'online-cmvn', (#3383)
KarelVesely84 Jun 10, 2019
c10e02f
[build,src] Enhancements to the cudamatrix/cudavector classes. (#3373)
luitjens Jun 11, 2019
b0a6e76
[egs] Fix perl `use encoding` deprecation (#3386)
naxingyu Jun 11, 2019
63c54e2
[scripts] Add max_active to align_fmllr_lats.sh to prevent rare crash…
hhadian Jun 11, 2019
7c7a176
[src] Implemented CUDA acclerated online cmvn. (#3370)
luitjens Jun 11, 2019
c7876a3
[egs] Fixed file path RE augmentation, in aspire recipe (#3388)
Shujian2015 Jun 12, 2019
0552e22
[scripts] Update taint_ctm_edits.py, RE utf-8 encoding (#3392)
vimalmanohar Jun 13, 2019
63b3849
[src] Change nnet3-am-copy to allow more manipulations (#3393)
danpovey Jun 14, 2019
c216385
[egs] Remove confusing setting of overridden num_epochs variable in a…
Shujian2015 Jun 15, 2019
bd1da14
[build] Add a missing dependency for "decoder" in Makefile (#3397)
hhadian Jun 17, 2019
674410e
[src] CUDA decoder performance patch (#3391)
hugovbraun Jun 19, 2019
10f2fcb
[build,scripts] Dependency fix; add cross-references to scripts (#3400)
danpovey Jun 19, 2019
76557e9
[egs] Fix cleanup-after-partial-download bug in aishell (#3404)
naxingyu Jun 20, 2019
5abb1a0
[src] Change functions like AppiyLog() to all work out-of-place (#3185)
YiwenShaoStephen Jun 20, 2019
777f8c1
[src] Make stack trace display more user friendly (#3406)
rosun82 Jun 21, 2019
d5a1451
[egs] Fix to separators in Aspire reverb recipe (#3408)
danpovey Jun 22, 2019
09697c3
[egs] Fix to separators in Aspire, related to #3408 (#3409)
Shujian2015 Jun 22, 2019
fe541d2
[src] online2-tcp, add option to display start/end times (#3399)
KarelVesely84 Jun 23, 2019
c5c09e9
[src] Remove debugging assert in cuda feature extraction code (#3411)
luitjens Jun 24, 2019
837839a
[scripts] Fix to checks in adjust_unk_graph.sh (#3410)
hhadian Jun 24, 2019
563b258
[src] Added GPU feature extraction (will improve speed of GPU decodin…
luitjens Jun 24, 2019
ec13b71
[src] Fix build error introducted by race condition in PR requests/ac…
luitjens Jun 24, 2019
5b4d2c9
[src] Added error string to CUDA allocation errors. (#3413)
luitjens Jun 25, 2019
00963e2
[src] Fix CUDA_VSERION number in preprocessor checks (#3414)
LvHang Jun 25, 2019
14cc156
[src] Fix build of online feature extraction with older CUDA version …
luitjens Jun 26, 2019
5cc7ce0
[src] Update Insert function of hashlist and decoders (#3402)
LvHang Jun 26, 2019
524db19
[src] Fix spelling mistake in #3415 (#3416)
LvHang Jun 26, 2019
36a7e99
[build] Fix configure bug RE CuSolver (#3417)
LvHang Jun 26, 2019
533469c
[src] Enable an option to use the GPU for feature extraction in GPU d…
luitjens Jun 26, 2019
42315f3
[egs] Replace $cuda_cmd with $train_cmd for FarsDat (#3426)
rezame Jun 27, 2019
31df26c
[src] Remove outdated comment (#3148) (#3422)
cloudhan Jun 27, 2019
2e02eb7
[src] Adding missing thread.join in CUDA decoder and fixing two todos…
hugovbraun Jun 27, 2019
f5a5f84
[build] Add missing lib dependency in cudafeatbin (#3427)
danpovey Jun 27, 2019
fa2e8c3
[egs] Small fix to aspire run_tdnn_7b.sh (#3429)
Shujian2015 Jun 28, 2019
21c0d9b
[build] Fix to cuda makefiles, thanks: yiyidhuang@gmail.com (#3431)
danpovey Jun 28, 2019
f5d34d7
[build] Add missing deps to cuda makefiles, thanks: yiyidhuang@gmail.…
danpovey Jun 28, 2019
8c0277e
[egs] Fix encoding issues in Chinese ASR recipe (#3430) (#3434)
boystray Jun 29, 2019
b7845dd
Revert "[src] Update Insert function of hashlist and decoders (#3402)…
danpovey Jun 29, 2019
0dcc2c9
[src] Update Insert function of hashlist and decoders (#3402) (#3438)
LvHang Jun 29, 2019
c449031
[build] Fix the cross-compiling issue for Android under MacOS (#3435)
rayworks Jun 30, 2019
9a38007
[src] Marking operator as __host__ __device__ to avoid build issues (…
luitjens Jul 2, 2019
ab4eca0
[egs] Fix perl encoding bug (was causing crashes) (#3442)
naxingyu Jul 2, 2019
893181f
[src] Cuda decoder fixes, efficiency improvements (#3443)
hugovbraun Jul 3, 2019
a4b6388
[scripts] Fix shebang of taint_ctm_edits.py to invoke python3 directl…
naxingyu Jul 3, 2019
e15f689
[src] Fix to a check in nnet-compute code (#3447)
danpovey Jul 3, 2019
f53556e
[src,scripts] Various typo fixes and stylistic fixes (#3153)
csukuangfj Jul 4, 2019
7800a0b
[scripts] Scripts for VB (variational bayes) resegmentation for Callh…
HuangZiliAndy Jul 7, 2019
85d1acd
[scripts] Extend utils/data/subsegment_data_dir.sh to copy reco2dur (…
yonatankahana Jul 8, 2019
1badfd9
[src,scripts,egs] Add code and example for SpecAugment in nnet3 (#3449)
danpovey Jul 9, 2019
39b7898
[scripts] Make segment_long_utterance honor frame_shift (#3455)
frankz-ai Jul 9, 2019
b5385b4
[scripts] Fix to steps/nnet/train.sh (nnet1) w.r.t. incorrect bash te…
xiaominfc Jul 10, 2019
98aa1d8
[egs] fixed bug in egs/gale_arabic/s5c/local/prepare_dict_subword.sh …
DongjiGao Jul 17, 2019
7664fa2
[src,build] Small fixes (#3472)
aka-rider Jul 19, 2019
abfb99d
[egs] Add MGB-2 Arabic recipe (#3333)
vimalmanohar Jul 19, 2019
29b3265
[scripts] Check/fix utt2num_frames when fixing data dir. (#3482)
hhadian Jul 20, 2019
4f4d3ff
[src] A couple small bug fixes. (#3477)
danpovey Jul 23, 2019
06442e1
[src,scripts] Cosmetic,file-mode fixes; fix to nnet1 align.sh introdu…
danpovey Jul 23, 2019
b876c28
[egs] Add missing script in MGB2 recipe (#3491)
vimalmanohar Jul 27, 2019
07c9d97
[egs] Fixing nnet1 but introduced in #3383 (rel. to #3487) (#3494)
danpovey Jul 27, 2019
7637de7
[src] Fix for nnet3 bug encountered when implementing deltas. (#3495)
danpovey Jul 28, 2019
ffda51e
[scripts,egs] Replace LDA layer with delta and delta-delta features (…
desh2608 Aug 2, 2019
cbae141
[egs] Add updated tdnn recipe for AMI (#3497)
aarora8 Aug 3, 2019
b0f0ae6
[egs] Create MALACH recipe based on s5b for AMI (#3496)
picheny Aug 3, 2019
5b2604a
[scripts] add --phone-symbol-table to prepare_lang_subword.sh (#3485)
DongjiGao Aug 4, 2019
0986892
[scripts] Option to prevent crash when adapting on much smaller data …
jtrmal Aug 5, 2019
53b6549
[build,scripts] Make OpenBLAS install check for gfortran; documentati…
danpovey Aug 5, 2019
fb5e7ac
[egs] Update chain TDNN-F recipe for CHIME s5 to match s5b, improves …
jtrmal Aug 5, 2019
b2b4ee8
[egs] Fix to #3505: updating chime5 TDNN-F script (#3508)
jtrmal Aug 5, 2019
b0be2e0
[scripts] Fixed issue that leads to empty segment file (#3510)
Shujian2015 Aug 6, 2019
83aeeac
[egs] Fix bug in AMI s5b RE overlapping segments that causes Fixed ov…
desh2608 Aug 8, 2019
5766a83
[egs] Small cosmetic change in extend_vocab_demo.sh (#3516)
QIU-Shuo Aug 12, 2019
97f0a45
[src] Cosmetic changes; fix windows-compile bug reported by @spencerk…
danpovey Aug 13, 2019
289edad
[src] Move cuda gpu from nnetbin to nnet3bin. (#3513)
danpovey Aug 13, 2019
1f6c760
fix a bug in egs/voxceleb/v1/local/make_voxceleb1_v2.pl when preparin…
jkchen79 Aug 13, 2019
662dc57
[egs] Fixed some bugs in mgb_data_prep.sh of mgb2_arabic (#3501)
DongjiGao Aug 13, 2019
1ff668a
[src,scripts] fix various typos and errors in comments (#3454)
csukuangfj Aug 13, 2019
998b220
[src] Move cuda-compiled to nnet3bin (#3517)
danpovey Aug 15, 2019
be26e62
[src] Fix binary name in Makefile, RE cuda-compiled (#3518)
danpovey Aug 15, 2019
3f95ed9
[src] buffer fix in cudafeat (#3521)
hugovbraun Aug 15, 2019
6cdc495
[src] Hopefully make it possible to use empty FST in grammar-fst (#3523)
danpovey Aug 19, 2019
317cad4
[src] Add option to convert pdf-posteriors to phones (#3526)
danpovey Aug 20, 2019
6c816e5
[src] Fix GetDeltaWeights for long-running online decoding (#3528)
nshmyrev Aug 22, 2019
0b7c9a9
[src] Small fix to post-to-phone-post.cc (problem introduced in from …
Aug 24, 2019
013fc50
[src]: adding Dan's fix to a bug in nnet-computation-graph (#3531)
david-ryan-snyder Aug 24, 2019
ca523c9
[egs] Replace prep_test_aspire_segmentation.sh (#2943) (#3530)
Shujian2015 Aug 24, 2019
93bdc41
[egs] OCR: Decomposition for CASIA and YOMDLE_ZH datasets (#3527)
ChunChiehChang Aug 24, 2019
080b5a2
[build] check_dependencies.sh: correct installation command for fedor…
103yiran Aug 26, 2019
6311057
[src,doc] Fix bug in new option of post-to-phone-post; skeleton of fa…
danpovey Aug 26, 2019
ed1d139
[egs,scripts] Adding possibility to have 'online-cmn' on input of 'n…
KarelVesely84 Aug 27, 2019
9f3d81f
[scripts] Fix to build_tree_multiple_sources.sh (#3545)
KimJeongSun Aug 28, 2019
605732a
[doc] Fix accidental overwriting of kws page (#3541)
danpovey Aug 28, 2019
f3b5f5b
[egs] Fix regex filter in Fisher preprocessing (was excluding 2-lette…
vineelpratap Aug 28, 2019
93d5baa
[scripts] Fix to bug introduced in #3498 RE ivector-extractor options…
danpovey Aug 28, 2019
745c0e8
[scripts] Fix awk compatibility issue; be more careful about online_c…
danpovey Aug 28, 2019
61bc12e
[src] Add a method for backward-compatibility with previous API (#3536)
nshmyrev Aug 28, 2019
9a83681
[src] Feature bank feature extraction using CUDA (#3544)
LeviBarnes Aug 28, 2019
66c9889
[src] Fix missing semicolon (#3551)
freewym Aug 29, 2019
fb5dbf4
[src] fix a typo mistaking not equal for assign sign in CUDA feature …
freewym Aug 29, 2019
8875a80
[src] Fix issue #3401 (crash in ivector extraction with max-remembere…
danpovey Aug 29, 2019
54e9a84
[egs] Librispeech: in RESULTS, change best_wer.sh => utils/best_wer.s…
astropeak Aug 29, 2019
72ca1eb
[egs] semisupervised recipes: fixing some variables in comments (#3547)
teinhonglo Aug 29, 2019
36525ba
[scripts] fix utils/lang/extend_lang.sh to add nonterm symbols in al…
lucas-jo Aug 30, 2019
fcea01a
[scripts] Fix to bug in steps/data/data_dir_manipulation_lib.py (#3174)
danpovey Sep 1, 2019
f3e77b1
[src] Fix in nnet3-attention-component.cc, RE required context (#3563)
danpovey Sep 3, 2019
8834e8b
[src] Temporarily turn off some model-collapsing code while investiga…
danpovey Sep 3, 2019
7c8e66e
[scripts] Fix to data cleanup script RE utf-8 support in uttearnce
sih4sing5hong5 Sep 6, 2019
09abdda
[src,scripts,egs] online-cmvn for online2 with chain models, (#3560)
KarelVesely84 Sep 7, 2019
ad09b19
[build,src] Change to error message; update kaldi_lm install script (…
danpovey Sep 10, 2019
1738fe6
[src] Clarify something in plda.h (cosmetic change) (#3588)
danpovey Sep 14, 2019
a74807b
[src] Small fix to online ivector extraction (RE #3401/#3405), thanks…
danpovey Sep 16, 2019
06887c8
[egs] Add recipe for Chinese training with multiple databases (#3555)
naxingyu Sep 16, 2019
2453b61
[src] Remove duplicate `num_done++` in apply-cmvn-online.cc (#3597)
ferb2015 Sep 17, 2019
cf1bcb6
[src] Fix to CMVN with CUDA (#3593)
LeviBarnes Sep 18, 2019
21bd4cf
[src] Fix two bugs in batched-wav-nnet3-cuda binary. (#3600)
luitjens Sep 19, 2019
fdab49d
[scripts] propagated bad whitespace fix to validate_dict_dir.pl (cosm…
pskrunner14 Sep 19, 2019
effe62e
[scripts] bug-fix on subset_data_dir.sh with --per-spk option (#3567)…
schaltung Sep 20, 2019
b2f64fe
[src] Code changes to support GCC9 + OpenFST1.7.3 + C++2a (namespace …
jtrmal Sep 22, 2019
8634823
[scripts] Training ivector-extractor: make online-cmvn per speaker. (…
KarelVesely84 Sep 25, 2019
fb2bd46
[src] cached compiler I/O for nnet3-xvector-compute (#3197)
stanleyguan Sep 25, 2019
7e52cf9
[src] Fixed online2-tcp-nnet3-decode-faster.cc poll_ret checks (#3611)
babakrkb Sep 29, 2019
cc62e17
[scripts] Call utils/data/get_utt2dur.sh using the correct $cmd and $…
alumae Sep 29, 2019
6b1ef76
[scripts] Enable tighter control of downloaded dependencies (#3543) (…
iskunk Sep 29, 2019
5d1fff2
[scripts] Make reverberate_data_dir.py handle vad.scp (#3619)
danpovey Sep 29, 2019
b10e474
[scripts] Don't get utt2dur in librispeech.. will be made by make_mfc…
danpovey Sep 29, 2019
924b146
[scripts] Make combine_ali_dirs.sh work when queue.pl is used (#3537)…
KimJeongSun Sep 29, 2019
a617bc4
[egs] Fix duplicate removal of unk from Librispeech decoding graphs (…
danpovey Sep 29, 2019
cebe114
[build] Check for gfortran, needed by OpenBLAS (for lapack) (#3622)
danpovey Sep 29, 2019
fc64b27
[scripts] VB resegmentation: load MFCCs only when used (save memory) …
HuangZiliAndy Sep 29, 2019
cb1b273
[build] removed old Docker files - see docker in the root folder for …
mdoulaty Sep 29, 2019
76b808a
[src] Fix to compute-mfcc-feats.cc (thanks: @puneetbawa) (#3623)
danpovey Sep 29, 2019
6ffde4b
[egs] Update AMI tdnn recipe to reflect what was really run, and add …
aarora8 Sep 29, 2019
8dcae07
[egs] Fix to run_ivector_common.sh in swbd (crash when running from s…
sih4sing5hong5 Oct 2, 2019
7b4c577
[scripts] Make data augmentation work with UTF-8 utterance ids/filena…
sih4sing5hong5 Oct 3, 2019
e1be9b8
[src] fix a bug in src/online/online-faster-decoder.cc (prevent segfa…
freewym Oct 3, 2019
41140d0
[scripts] Make extend_lang.sh support lexiconp_silprob.txt (#3339) (…
DongjiGao Oct 3, 2019
317e175
[scripts] Fix typo in analyze_alignments.sh (#3635)
mgoldey Oct 4, 2019
68cfbba
[scripts] Change the GPU policy of align.sh to wait instead of yes (#…
frankz-ai Oct 4, 2019
743eb23
[egs] Added tri3b and chain training for Aurora4 (#3638)
desh2608 Oct 4, 2019
526ade1
[build] fixed broken Docker builds by adding gfortran package (#3640)
mdoulaty Oct 5, 2019
b5f4cfc
[src] Fix bug in resampling checks for online features (#3639)
entn-at Oct 5, 2019
c2e0394
[build] Bump OpenBLAS version to 0.3.7 and enable locking (#3642)
sayint-ai Oct 8, 2019
8a1be79
[scripts] Update nnet3_to_dot.py to ignore Scale descriptors (#3644)
rouseabout Oct 9, 2019
a4de025
[src] Speed optimization for decoders (call templates properly) (#3637)
Bar-BY Oct 9, 2019
06bf62a
[doc] update FAQ page to include some FAQ candidates from Kaldi maili…
Oct 10, 2019
e601fda
[egs] Small fix: duplicate xent settings from examples (#3649)
sih4sing5hong5 Oct 11, 2019
e45dfb0
[doc] Fix typo (#3648)
sih4sing5hong5 Oct 11, 2019
e2b9180
[egs] Fix a bug in Tedlium run_ivector_common.sh (#3647)
jerrykuo7727 Oct 11, 2019
eeef080
[src] Fix to matrix-lib-test to ignore small difference (#3650)
danpovey Oct 12, 2019
3492d6e
[doc] update FAQ page: (#3651)
Oct 14, 2019
f8bf717
[scripts] Modify split_data.sh to split data evenly when utt2dur exis…
RuABraun Oct 15, 2019
2f1bca3
[doc] update FAQ page: added section for free dataset, python wrapper…
Oct 15, 2019
b39ed47
[src] Some CUDA i-vector fixes (#3660)
luitjens Oct 17, 2019
01cfda3
[src] CUDA batched decoder pipeline fixes (#3659)
luitjens Oct 17, 2019
b57ff17
[egs] Small fix to aspire example: pass in num_data_reps (#3665)
Shujian2015 Oct 18, 2019
880df12
[doc] Update FAQ page (#3663)
qindazhu Oct 18, 2019
ea85eed
[src] Add batched xvector computation (#3643)
danpovey Oct 19, 2019
be88f30
[src] CUDA decoder: write all output to single stream (#3666)
luitjens Oct 19, 2019
47d516d
[egs] Fix sox command in multi-Updated sox command (#3667)
Oct 20, 2019
7f0c9c1
[src] Change data-type for hash in lattice alignment (avoid debugger …
jiedanzhu2307 Oct 21, 2019
8151017
[egs] Two fixes to multi_en setup (#3676)
Oct 23, 2019
e6d5614
[scripts] Fix misspelled variable in reverb script (#3678)
Alicegaz Oct 23, 2019
a911208
Update cudamatrix.dox (#3674)
valboldakov Oct 24, 2019
eb83683
[scripts] Add reduced-context option for TDNN-F layers. (#3658)
danpovey Oct 24, 2019
e2c006c
[egs] Make librispeech run_tdnn_discriminative.sh more efficient with…
jiangj-dc Oct 26, 2019
e3a1b99
[egs] Remove pitch from multi_cn nnet3 recipe (#3686)
naxingyu Oct 28, 2019
9357553
[egs] multi_cn: clarify metric is CER (#3687)
naxingyu Oct 28, 2019
d91beea
[src,minor] Fix typo in comment in kaldi-holder.h (#3691)
zh794390558 Oct 29, 2019
d649afd
[egs] update run_tdnn_discriminative.sh in librispeech recipe (#3689)
jiangj-dc Oct 29, 2019
03dccb2
[egs] Aspire recipe: fixed typo in utterance and speaker prefix (#3696)
skolachi Oct 31, 2019
7249cc0
[src] cuda batched decoder, fix memory bugs (#3697)
luitjens Nov 1, 2019
f21d7e7
[scripts] Change make_rttm.py to read/write files with UTF-8 encoding…
entn-at Nov 7, 2019
c2a5538
[src] Removing non-compiling paranoid asserts in nnet-computation-gra…
RuABraun Nov 9, 2019
bf0ee72
[build] Fix gfortran package name for centos (#3708)
rezame Nov 9, 2019
f1dd9e2
[scripts] Change the Python diagnostic scripts to accept non-ASCII UT…
vdp Nov 11, 2019
45cae5a
[scripts] Fix some issues in #3653 in split_scp.pl (#3710)
RuABraun Nov 12, 2019
898c8da
[scripts] Fix 2 issues in nnet2->nnet3 model conversion script (#886)…
jfainberg Nov 12, 2019
f679c78
Removing changes to split_scp.pl (#3717)
danpovey Nov 14, 2019
ab36598
[scripts] Improve how combine_ali_dirs.sh gets job-specific filenames…
KimJeongSun Nov 15, 2019
d97f1d8
[src] Add --debug-level=N configure option to control assertions (#36…
iskunk Nov 16, 2019
0cc71f1
[src] Adding some more feature extraction options (needed by some use…
jtrmal Nov 21, 2019
249e2b2
[src,script,egs] Goodness of Pronunciation (GOP) (#3703)
jimbozhang Nov 21, 2019
59255ae
[src] Making ivector extractor tolerate dim mismatch due to pitch (#3…
danpovey Nov 21, 2019
eb28a6a
Revert "[src] Making ivector extractor tolerate dim mismatch due to p…
danpovey Nov 21, 2019
bbbc79b
[src] Fix NVCC compilation errors on Windows (#3741)
cloudhan Nov 28, 2019
f88c475
[build] Add CMake Build System as alternative to current Makefile-bas…
cloudhan Nov 28, 2019
1d0b267
[scripts] Modify split_data_dir.sh and split_scp.pl to use utt2dur if…
desh2608 Nov 30, 2019
915bb78
[scripts] fix slurm.pl error (#3745)
naxingyu Nov 30, 2019
666b8cb
Revert "[scripts] Modify split_data_dir.sh and split_scp.pl to use ut…
danpovey Dec 1, 2019
413c7c8
[egs] Children's speech ASR recipe for cmu_kids and cslu_kids (#3699)
Fei00Wu Dec 2, 2019
1cd7ee9
[src] Incremental determinization [cleaned up/rewrite] (#3737)
danpovey Dec 2, 2019
d77457d
[scripts] Add scripts to create combine fmllr-tranform dirs(#3752)
marvin-nj Dec 3, 2019
018d180
[src] CUDA decoder: fix invalid-lattice error that happens in corner …
hugovbraun Dec 5, 2019
be2dbf4
[egs] Add Chime 6 baseline system (#3755)
aarora8 Dec 5, 2019
daf9d6e
[scripts] Fix issue in copy_lat_dir.sh affecting combine_lat_dirs.sh …
marvin-nj Dec 5, 2019
07d02da
[src] Add missing #include, needed for CUDA decoder compilation on so…
luitjens Dec 6, 2019
6f329a6
[scripts] fix bug in steps/data/reverberate_data_dir.py (#3762)
HuangZiliAndy Dec 8, 2019
d0007f3
[src] CUDA allocator: fix order of next largest block (#3739)
btiplitz Dec 11, 2019
454fb08
[egs] some fixes in Chime6 baseline system (#3763)
aarora8 Dec 11, 2019
42c3888
[src] CUDA decoding: add support for affine transforms to CUDA featur…
luitjens Dec 11, 2019
5ca36b9
[src] relax assertion constraint slightly (RE matrix orthonormalizati…
freewym Dec 11, 2019
cba392c
[src] CUDA decoder: fix bug in NumPendingTasks() (#3769)
luitjens Dec 12, 2019
ea5757a
[src] Add options to select specific gpu, reuse cuda context (#3750)
hugovbraun Dec 12, 2019
799e1f0
[src] Move CheckAndFix to config struct (#3749)
hugovbraun Dec 13, 2019
e27bbda
[egs,scripts] Add recipes for CN-Celeb (#3758)
Dec 14, 2019
1be4750
[src] CUDA decoder: remove unecessary sync that was added for debuggi…
luitjens Dec 14, 2019
fe7b922
[src] CUDA decoder: shrink channel vectors instead of vector holding …
luitjens Dec 17, 2019
9582e71
Merge remote-tracking branch 'origin/master' into HEAD
naxingyu Dec 17, 2019
255c666
add include path
naxingyu Dec 17, 2019
eadcf1f
update test
naxingyu Dec 17, 2019
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
18 changes: 18 additions & 0 deletions .github/ISSUE_TEMPLATE/bug_report.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
---
name: Bug report
about: Create a report to help us improve
title: ''
labels: bug
assignees: ''

---

<!--
WARNING: THE KALDI ISSUE TRACKER IS **ONLY** USED FOR KALDI DEVELOPMENT!

If you have a question about using Kaldi, please use the kald-help discussion group:

https://groups.google.com/forum/#!forum/kaldi-help

Instructions for joining are available at: http://kaldi-asr.org/forums.html
-->
18 changes: 18 additions & 0 deletions .github/ISSUE_TEMPLATE/feature-proposal-discussion.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
---
name: Feature proposal or discussion
about: Suggest an idea for Kaldi
title: ''
labels: discussion
assignees: ''

---

<!--
WARNING: THE KALDI ISSUE TRACKER IS **ONLY** USED FOR KALDI DEVELOPMENT!

If you have a question about using Kaldi, please use the kald-help discussion group:

https://groups.google.com/forum/#!forum/kaldi-help

Instructions for joining are available at: http://kaldi-asr.org/forums.html
-->
21 changes: 17 additions & 4 deletions .gitignore
Original file line number Diff line number Diff line change
Expand Up @@ -73,15 +73,19 @@ GSYMS
/src/kaldi.mk.bak

# /egs/
/egs/*/s*/mfcc
/egs/*/s*/plp
/egs/*/s*/exp
/egs/*/s*/data
/egs/*/*/mfcc
/egs/*/*/plp
/egs/*/*/exp
/egs/*/*/data
/egs/*/*/wav
/egs/*/*/enhan

# /tools/
/tools/pocolm/
/tools/ATLAS/
/tools/atlas3.8.3.tar.gz
/tools/irstlm/
/tools/mitlm/
/tools/openfst
/tools/openfst-1.3.2.tar.gz
/tools/openfst-1.3.2/
Expand Down Expand Up @@ -143,3 +147,12 @@ GSYMS
/tools/mmseg-1.3.0.tar.gz
/tools/mmseg-1.3.0/
/kaldiwin_vs*
/tools/cub-1.8.0.zip
/tools/cub-1.8.0/
/tools/cub
/tools/python/

# These CMakeLists.txt files are all genareted on the fly at the moment.
# They are added here to avoid accidently checkin.
/src/**/CMakeLists.txt
/build*
2 changes: 1 addition & 1 deletion .travis.yml
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ script:
# for the explanation why extra switches needed for clang with ccache.
- CXX="ccache clang++-3.8 -Qunused-arguments -fcolor-diagnostics -Wno-tautological-compare"
CFLAGS=""
LDFLAGS="-llapack"
LDFLAGS="-llapack -Wl,-fuse-ld=gold"
INCDIRS="$XROOT/usr/include"
LIBDIRS="$XROOT/usr/lib"
tools/extras/travis_script.sh
Expand Down
195 changes: 195 additions & 0 deletions CMakeLists.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,195 @@
cmake_minimum_required(VERSION 3.5)
project(kaldi)

set(CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake;${CMAKE_MODULE_PATH}")
include(GNUInstallDirs)
include(Utils)
include(third_party/get_third_party)

message(STATUS "Running gen_cmake_skeleton.py")
execute_process(COMMAND python
"${CMAKE_CURRENT_SOURCE_DIR}/cmake/gen_cmake_skeleton.py"
"${CMAKE_CURRENT_SOURCE_DIR}/src"
"--quiet"
)

set(CMAKE_CXX_STANDARD 14)
set(CMAKE_CXX_EXTENSIONS OFF)
set(CMAKE_INSTALL_MESSAGE LAZY) # hide "-- Up-to-date: ..."
if(BUILD_SHARED_LIBS)
set(CMAKE_POSITION_INDEPENDENT_CODE ON)
if(WIN32)
set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
message(FATAL_ERROR "DLL is not supported currently")
elseif(APPLE)
set(CMAKE_INSTALL_RPATH "@loader_path")
else()
set(CMAKE_INSTALL_RPATH "$ORIGIN;$ORIGIN/../lib")
endif()
endif()

set(MATHLIB "OpenBLAS" CACHE STRING "OpenBLAS|MKL|Accelerate")
option(KALDI_BUILD_EXE "If disabled, will make add_kaldi_executable a no-op" ON)
option(KALDI_BUILD_TEST "If disabled, will make add_kaldi_test_executable a no-op" ON)
option(KALDI_USE_PATCH_NUMBER "Use MAJOR.MINOR.PATCH format, otherwise MAJOR.MINOR" OFF)

link_libraries(${CMAKE_DL_LIBS})

find_package(Threads)
link_libraries(Threads::Threads)

if(MATHLIB STREQUAL "OpenBLAS")
set(BLA_VENDOR "OpenBLAS")
find_package(LAPACK REQUIRED)
add_definitions(-DHAVE_CLAPACK=1)
include_directories(${CMAKE_CURRENT_SOURCE_DIR}/tools/CLAPACK)
link_libraries(${BLAS_LIBRARIES} ${LAPACK_LIBRARIES})
elseif(MATHLIB STREQUAL "MKL")
set(BLA_VENDOR "Intel10_64lp")
# find_package(BLAS REQUIRED)
normalize_env_path(ENV{MKLROOT})
find_package(LAPACK REQUIRED)
add_definitions(-DHAVE_MKL=1)
include_directories($ENV{MKLROOT}/include) # TODO: maybe not use env, idk, find_package doesnt handle includes...
link_libraries(${BLAS_LIBRARIES} ${LAPACK_LIBRARIES})
elseif(MATHLIB STREQUAL "Accelerate")
set(BLA_VENDOR "Apple")
find_package(BLAS REQUIRED)
find_package(LAPACK REQUIRED)
add_definitions(-DHAVE_CLAPACK=1)
link_libraries(${BLAS_LIBRARIES} ${LAPACK_LIBRARIES})
else()
message(FATAL_ERROR "${MATHLIB} is not tested and supported, you are on your own now.")
endif()

if(MSVC)
# Added in source, but we actually should do it in build script, whatever...
# add_definitions(-DWIN32_LEAN_AND_MEAN=1)

add_compile_options(/permissive- /FS /wd4819 /EHsc /bigobj)

# some warnings related with fst
add_compile_options(/wd4018 /wd4244 /wd4267 /wd4291 /wd4305)

set(CUDA_USE_STATIC_CUDA_RUNTIME OFF CACHE INTERNAL "")
if(NOT DEFINED ENV{CUDAHOSTCXX})
set(ENV{CUDAHOSTCXX} ${CMAKE_CXX_COMPILER})
endif()
if(NOT DEFINED CUDA_HOST_COMPILER)
set(CUDA_HOST_COMPILER ${CMAKE_CXX_COMPILER})
endif()
endif()

find_package(CUDA)
if(CUDA_FOUND)
set(CUB_ROOT_DIR "${PROJECT_SOURCE_DIR}/tools/cub")

set(CUDA_PROPAGATE_HOST_FLAGS ON)
set(KALDI_CUDA_NVCC_FLAGS "--default-stream=per-thread;-std=c++${CMAKE_CXX_STANDARD}")
if(MSVC)
list(APPEND KALDI_CUDA_NVCC_FLAGS "-Xcompiler /permissive-,/FS,/wd4819,/EHsc,/bigobj")
list(APPEND KALDI_CUDA_NVCC_FLAGS "-Xcompiler /wd4018,/wd4244,/wd4267,/wd4291,/wd4305")
if(BUILD_SHARED_LIBS)
list(APPEND CUDA_NVCC_FLAGS_RELEASE -Xcompiler /MD)
list(APPEND CUDA_NVCC_FLAGS_DEBUG -Xcompiler /MDd)
endif()
else()
# list(APPEND KALDI_CUDA_NVCC_FLAGS "-Xcompiler -std=c++${CMAKE_CXX_STANDARD}")
list(APPEND KALDI_CUDA_NVCC_FLAGS "-Xcompiler -fPIC")
endif()
set(CUDA_NVCC_FLAGS ${KALDI_CUDA_NVCC_FLAGS} ${CUDA_NVCC_FLAGS})

add_definitions(-DHAVE_CUDA=1)
add_definitions(-DCUDA_API_PER_THREAD_DEFAULT_STREAM=1)
include_directories(${CUDA_INCLUDE_DIRS})
link_libraries(
${CUDA_LIBRARIES}
${CUDA_CUDA_LIBRARY}
${CUDA_CUBLAS_LIBRARIES}
${CUDA_CUFFT_LIBRARIES}
${CUDA_curand_LIBRARY}
${CUDA_cusolver_LIBRARY}
${CUDA_cusparse_LIBRARY})

find_package(NvToolExt REQUIRED)
include_directories(${NvToolExt_INCLUDE_DIR})
link_libraries(${NvToolExt_LIBRARIES})

find_package(CUB REQUIRED)
include_directories(${CUB_INCLUDE_DIR})
endif()

add_definitions(-DKALDI_NO_PORTAUDIO=1)

include(VersionHelper)
get_version() # this will set KALDI_VERSION and KALDI_PATCH_NUMBER
if(${KALDI_USE_PATCH_NUMBER})
set(KALDI_VERSION "${KALDI_VERSION}.${KALDI_PATCH_NUMBER}")
endif()

get_third_party(openfst)
set(OPENFST_ROOT_DIR ${CMAKE_CURRENT_BINARY_DIR}/openfst)
include(third_party/openfst_lib_target)
link_libraries(fst)

# add all native libraries
add_subdirectory(src/base) # NOTE, we need to patch the target with version from outside
set_property(TARGET kaldi-base PROPERTY COMPILE_DEFINITIONS "KALDI_VERSION=\"${KALDI_VERSION}\"")
add_subdirectory(src/matrix)
add_subdirectory(src/cudamatrix)
add_subdirectory(src/util)
add_subdirectory(src/feat)
add_subdirectory(src/tree)
add_subdirectory(src/gmm)
add_subdirectory(src/transform)
add_subdirectory(src/sgmm2)
add_subdirectory(src/fstext)
add_subdirectory(src/hmm)
add_subdirectory(src/lm)
add_subdirectory(src/decoder)
add_subdirectory(src/lat)
add_subdirectory(src/nnet)
add_subdirectory(src/nnet2)
add_subdirectory(src/nnet3)
add_subdirectory(src/rnnlm)
add_subdirectory(src/chain)
add_subdirectory(src/ivector)
add_subdirectory(src/online)
add_subdirectory(src/online2)
add_subdirectory(src/kws)

add_subdirectory(src/itf)

# add all cuda libraries
if(CUDA_FOUND)
add_subdirectory(src/cudafeat)
add_subdirectory(src/cudadecoder)
endif()

# add all native executables
add_subdirectory(src/gmmbin)
add_subdirectory(src/featbin)
add_subdirectory(src/onlinebin)

# add all cuda executables
if(CUDA_FOUND)
add_subdirectory(src/cudafeatbin)
add_subdirectory(src/cudadecoderbin)
endif()

include(CMakePackageConfigHelpers)
# maybe we should put this into subfolder?
configure_package_config_file(
${CMAKE_CURRENT_SOURCE_DIR}/cmake/kaldi-config.cmake.in
${CMAKE_BINARY_DIR}/cmake/kaldi-config.cmake
INSTALL_DESTINATION lib/cmake/kaldi
)
write_basic_package_version_file(
${CMAKE_BINARY_DIR}/cmake/kaldi-config-version.cmake
VERSION ${KALDI_VERSION}
COMPATIBILITY AnyNewerVersion
)
install(FILES ${CMAKE_BINARY_DIR}/cmake/kaldi-config.cmake ${CMAKE_BINARY_DIR}/cmake/kaldi-config-version.cmake
DESTINATION lib/cmake/kaldi
)
install(EXPORT kaldi-targets DESTINATION ${CMAKE_INSTALL_PREFIX}/lib/cmake/kaldi)
17 changes: 12 additions & 5 deletions INSTALL
Original file line number Diff line number Diff line change
@@ -1,9 +1,16 @@
This is the official Kaldi INSTALL. Look also at INSTALL.md for the git mirror installation.
[for native Windows install, see windows/INSTALL]
[Option 1 in the following does not apply to native Windows install, see windows/INSTALL or following Option 2]

(1)
go to tools/ and follow INSTALL instructions there.
Option 1 (bash + makefile):

(2)
go to src/ and follow INSTALL instructions there.
Steps:
(1)
go to tools/ and follow INSTALL instructions there.

(2)
go to src/ and follow INSTALL instructions there.

Option 2 (cmake):

Go to cmake/ and follow INSTALL.md instructions there.
Note, it may not be well tested and some features are missing currently.
Loading