Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update README.md in egs2/librimix/tse1 and egs2/wsj0_2mix/tse1 #5289

Merged
merged 6 commits into from
Jul 21, 2023
Merged

Update README.md in egs2/librimix/tse1 and egs2/wsj0_2mix/tse1 #5289

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
ff3cf3c
Update README.md in egs2/librimix/tse1 and egs2/wsj0_2mix/tse1
Emrys365 Jul 9, 2023
f680b8c
Update README.md in egs2/librimix/tse1 and egs2/wsj0_2mix/tse1
Emrys365 Jul 9, 2023
72a0097
Update README.md in egs2/wsj0_2mix/tse1
Emrys365 Jul 9, 2023
8def626
Fix a bug in transcript preparation in egs2/wsj0_2mix/tse1
Emrys365 Jul 15, 2023
05aba16
Merge branch 'master' of github.com:espnet/espnet into tse
Emrys365 Jul 21, 2023
233e903
Merge branch 'master' into tse
sw005320 Jul 21, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
22 changes: 22 additions & 0 deletions egs2/librimix/tse1/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,22 @@
<!-- Generated by ./scripts/utils/show_enh_score.sh -->
# RESULTS
## Environments
- date: `Mon Jun 5 22:42:07 CST 2023`
- python version: `3.8.16 (default, Mar 2 2023, 03:21:46) [GCC 11.2.0]`
- espnet version: `espnet 202301`
- pytorch version: `pytorch 2.0.1`
- Git hash: ``
- Commit date: ``


## enh_train_raw

- config: ./conf/train.yaml
- Pretrained model: https://huggingface.co/espnet/Wangyou_Zhang_librimix_train_enh_tse_td_speakerbeam_raw

|dataset|PESQ_WB|STOI|SAR|SDR|SIR|SI_SNR|OVRL|SIG|BAK|P808_MOS|
|---|---|---|---|---|---|---|---|---|---|---|
|dev|1.08|64.43|7.18|-1.71|0.08|-1.81|1.60|2.26|1.62|2.68|
|test|1.08|64.56|6.90|-1.83|0.09|-1.93|1.63|2.33|1.66|2.71|
|enhanced_dev|1.73|86.50|12.50|11.40|24.83|10.58|2.95|3.24|3.92|3.23|
|enhanced_test|1.73|87.36|12.34|11.47|24.51|10.74|2.99|3.29|3.91|3.25|
85 changes: 0 additions & 85 deletions egs2/librimix/tse1/conf/tuning/train_enh_tse_td_speakerbeam_4gpu_max.yaml
View file
Open in desktop

This file was deleted.

20 changes: 20 additions & 0 deletions egs2/wsj0_2mix/tse1/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
<!-- Generated by ./scripts/utils/show_enh_score.sh -->
# RESULTS
## Environments
- date: `Sun Jul 9 09:23:16 CST 2023`
- python version: `3.8.16 (default, Mar 2 2023, 03:21:46) [GCC 11.2.0]`
- espnet version: `espnet 202301`
- pytorch version: `pytorch 2.0.1`
- Git hash: ``
- Commit date: ``


## enh_train_enh_tse_td_speakerbeam_raw

- config: conf/tuning/train_enh_tse_td_speakerbeam.yaml
- Pretrained model: https://huggingface.co/espnet/Wangyou_Zhang_wsj0_2mix_train_enh_tse_td_speakerbeam_raw

|dataset|PESQ_NB|STOI|SAR|SDR|SIR|SI_SNR|
|---|---|---|---|---|---|---|
|enhanced_cv_min_8k|3.54|96.41|18.75|18.75|0.00|18.37|
|enhanced_tt_min_8k|3.46|96.35|17.51|17.51|0.00|17.11|
6 changes: 3 additions & 3 deletions egs2/wsj0_2mix/tse1/conf/tuning/train_enh_tse_td_speakerbeam.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ max_epoch: 100
batch_type: folded
batch_size: 4
iterator_type: chunk
chunk_length: 24000
chunk_length: 32000
# exclude keys "enroll_ref", "enroll_ref1", "enroll_ref2", ...
# from the length consistency check in ChunkIterFactory
chunk_excluded_key_prefixes:
Expand All @@ -23,7 +23,7 @@ val_scheduler_criterion:
- loss
best_model_criterion:
- - valid
- si_snr
- snr
- max
- - valid
- loss
Expand All @@ -40,7 +40,7 @@ model_conf:
share_encoder: true # this must be False if load_spk_embedding is True

train_spk2enroll: null
enroll_segment: 24000
enroll_segment: 16000
load_spk_embedding: false
load_all_speakers: false

Expand Down
6 changes: 3 additions & 3 deletions egs2/wsj0_2mix/tse1/local/wsj0_2mix_data_prep.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -121,8 +121,8 @@ done
# change to the original path
cd ..

awk '(ARGIND==1) {txt[$1]=$0} (ARGIND==2) {split($1, lst, "_"); if(FNR %2 != 0){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_tr_s.txt ${data}/${tr}/wav.scp | awk '{$2=""; print $0}' > ${data}/${tr}/text
awk '(ARGIND==1) {txt[$1]=$0} (ARGIND==2) {split($1, lst, "_"); if(FNR %2 != 0){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_tr_s.txt ${data}/${cv}/wav.scp | awk '{$2=""; print $0}' > ${data}/${cv}/text
awk '(ARGIND<=2) {txt[$1]=$0} (ARGIND==3) {split($1, lst, "_"); if(FNR %2 != 0){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_dt_05.txt tmp/si_et_05.txt ${data}/${tt}/wav.scp | awk '{$2=""; print $0}' > ${data}/${tt}/text
awk '(ARGIND==1) {txt[$1]=$0} (ARGIND==2) {n=split($1, lst, "_"); if(substr(lst[n],0,3) == substr(lst[3],0,3)){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_tr_s.txt ${data}/${tr}/wav.scp | awk '{$2=""; print $0}' > ${data}/${tr}/text
awk '(ARGIND==1) {txt[$1]=$0} (ARGIND==2) {n=split($1, lst, "_"); if(substr(lst[n],0,3) == substr(lst[3],0,3)){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_tr_s.txt ${data}/${cv}/wav.scp | awk '{$2=""; print $0}' > ${data}/${cv}/text
awk '(ARGIND<=2) {txt[$1]=$0} (ARGIND==3) {n=split($1, lst, "_"); if(substr(lst[n],0,3) == substr(lst[3],0,3)){utt1=lst[3];}else{utt1=lst[5];} text=txt[utt1]; print($1, text)}' tmp/si_dt_05.txt tmp/si_et_05.txt ${data}/${tt}/wav.scp | awk '{$2=""; print $0}' > ${data}/${tt}/text

rm -r tmp