mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-08-27 02:34:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

add speed perturb for kespeech

Browse Source
This commit is contained in:
Yuekai Zhang 2024-02-23 09:54:26 +08:00
parent 5a62723f19
commit 73e5caecc5
4 changed files with 17 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
egs/aishell4/ASR/prepare.sh
View File

@ -78,9 +78,9 @@ fi
if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then
log "Stage 2: Compute fbank for aishell4" log "Stage 2: Compute fbank for aishell4"
if [ ! -f data/fbank/aishell4/.fbank.done ]; then if [ ! -f data/fbank/aishell4/.fbank.done ]; then
mkdir -p data/fbank/aishell4 mkdir -p data/fbank
./local/compute_fbank_aishell4.py --perturb-speed ${perturb_speed} ./local/compute_fbank_aishell4.py --perturb-speed ${perturb_speed}
touch data/fbank/aishell4/.fbank.done touch data/fbank/.fbank.done
fi fi
fi fi
@ -88,9 +88,9 @@ whisper_mel_bins=80
if [ $stage -le 20 ] && [ $stop_stage -ge 20 ]; then if [ $stage -le 20 ] && [ $stop_stage -ge 20 ]; then
log "Stage 20: Compute whisper fbank for aishell4" log "Stage 20: Compute whisper fbank for aishell4"
if [ ! -f data/fbank/aishell4/.fbank.done ]; then if [ ! -f data/fbank/aishell4/.fbank.done ]; then
mkdir -p data/fbank/aishell4 mkdir -p data/fbank
./local/compute_fbank_aishell4.py --perturb-speed ${perturb_speed} --num-mel-bins ${whisper_mel_bins} --whisper-fbank true ./local/compute_fbank_aishell4.py --perturb-speed ${perturb_speed} --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
touch data/fbank/aishell4/.fbank.done touch data/fbank/.fbank.done
fi fi
fi fi

2
egs/alimeeting/ASR/local/compute_fbank_alimeeting.py
View File

@ -53,7 +53,7 @@ def compute_fbank_alimeeting(num_mel_bins: int = 80, perturb_speed: bool = False
"test", "test",
) )
prefix = "alimeeting" prefix = "alimeeting-far"
suffix = "jsonl.gz" suffix = "jsonl.gz"
manifests = read_manifests_if_cached( manifests = read_manifests_if_cached(
dataset_parts=dataset_parts, dataset_parts=dataset_parts,

10
egs/alimeeting/ASR/prepare.sh
View File

@ -67,18 +67,20 @@ fi
if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then
log "Stage 2: compute fbank for alimeeting" log "Stage 2: compute fbank for alimeeting"
if [ ! -f data/fbank/alimeeting/.fbank.done ]; then if [ ! -f data/fbank/.fbank.done ]; then
mkdir -p data/fbank/alimeeting mkdir -p data/fbank
./local/compute_fbank_alimeeting.py --perturb-speed ${perturb_speed} ./local/compute_fbank_alimeeting.py --perturb-speed ${perturb_speed}
touch data/fbank/.fbank.done
fi fi
fi fi
whisper_mel_bins=80 whisper_mel_bins=80
if [ $stage -le 20 ] && [ $stop_stage -ge 20 ]; then if [ $stage -le 20 ] && [ $stop_stage -ge 20 ]; then
log "Stage 20: compute whisper fbank for alimeeting" log "Stage 20: compute whisper fbank for alimeeting"
if [ ! -f data/fbank/alimeeting/.fbank.done ]; then if [ ! -f data/fbank/.fbank.done ]; then
mkdir -p data/fbank/alimeeting mkdir -p data/fbank
./local/compute_fbank_alimeeting.py --perturb-speed ${perturb_speed} --num-mel-bins ${whisper_mel_bins} --whisper-fbank true ./local/compute_fbank_alimeeting.py --perturb-speed ${perturb_speed} --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
touch data/fbank/.fbank.done
fi fi
fi fi

20
egs/multi_zh-hans/ASR/prepare.sh
View File

@ -5,8 +5,8 @@ export PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python
set -eou pipefail set -eou pipefail
stage=121 stage=120
stop_stage=121 stop_stage=120
num_splits=100 num_splits=100
dl_dir=$PWD/download dl_dir=$PWD/download
@ -86,7 +86,7 @@ fi
log "Dataset: AISHELL-2" log "Dataset: AISHELL-2"
if [ $stage -le 4 ] && [ $stop_stage -ge 4 ]; then if [ $stage -le 4 ] && [ $stop_stage -ge 4 ]; then
log "Stage 4: Prepare AISHELL-2" log "Stage 4: Prepare AISHELL-2"
if [ -e ../../aishell/ASR/data/fbank/.aishell2.done ]; then if [ -e ../../aishell2/ASR/data/fbank/.aishell2.done ]; then
cd data/fbank cd data/fbank
ln -svf $(realpath ../../../../aishell2/ASR/data/fbank/aishell2_feats_train) . ln -svf $(realpath ../../../../aishell2/ASR/data/fbank/aishell2_feats_train) .
ln -svf $(realpath ../../../../aishell2/ASR/data/fbank/aishell2_feats_dev) . ln -svf $(realpath ../../../../aishell2/ASR/data/fbank/aishell2_feats_dev) .
@ -104,7 +104,7 @@ fi
log "Dataset: AISHELL-4" log "Dataset: AISHELL-4"
if [ $stage -le 5 ] && [ $stop_stage -ge 5 ]; then if [ $stage -le 5 ] && [ $stop_stage -ge 5 ]; then
log "Stage 5: Prepare AISHELL-4" log "Stage 5: Prepare AISHELL-4"
if [ -e ../../aishell/ASR/data/fbank/.aishell4.done ]; then if [ -e ../../aishell4/ASR/data/fbank/.fbank.done ]; then
cd data/fbank cd data/fbank
ln -svf $(realpath ../../../../aishell4/ASR/data/fbank/aishell4_feats_train) . ln -svf $(realpath ../../../../aishell4/ASR/data/fbank/aishell4_feats_train) .
ln -svf $(realpath ../../../../aishell4/ASR/data/fbank/aishell4_feats_dev) . ln -svf $(realpath ../../../../aishell4/ASR/data/fbank/aishell4_feats_dev) .
@ -323,7 +323,7 @@ if [ $stage -le 120 ] && [ $stop_stage -ge 120 ]; then
log "Preprocess KeSpeech manifest" log "Preprocess KeSpeech manifest"
if [ ! -f data/fbank/.kespeech_preprocess_complete ]; then if [ ! -f data/fbank/.kespeech_preprocess_complete ]; then
python3 ./local/preprocess_kespeech.py python3 ./local/preprocess_kespeech.py --speed-perturb true
touch data/fbank/.kespeech_preprocess_complete touch data/fbank/.kespeech_preprocess_complete
fi fi
@ -350,20 +350,12 @@ if [ $stage -le 120 ] && [ $stop_stage -ge 120 ]; then
./local/compute_fbank_kespeech_splits.py --speed-perturb true --num-splits ${num_splits} --training-subset train_phase2 --num-mel-bins ${whisper_mel_bins} --whisper-fbank true ./local/compute_fbank_kespeech_splits.py --speed-perturb true --num-splits ${num_splits} --training-subset train_phase2 --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
log "Compute KeSpeech fbank for test/dev" log "Compute KeSpeech fbank for test/dev"
./local/compute_fbank_kespeech_dev_test.py --num-mel-bins ${whisper_mel_bins} --whisper-fbank true # ./local/compute_fbank_kespeech_dev_test.py --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
touch data/fbank/.kespeech.done touch data/fbank/.kespeech.done
fi fi
fi fi
if [ $stage -le 122 ] && [ $stop_stage -ge 122 ]; then
log "Stage 122: Prepare speed perturb versionKeSpeech for whisper"
./local/compute_fbank_kespeech_splits.py --speed-perturb true --num-splits ${num_splits} --training-subset train_phase1 --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
log "Compute KeSpeech fbank for train_phase2"
./local/compute_fbank_kespeech_splits.py --speed-perturb true --num-splits ${num_splits} --training-subset train_phase2 --num-mel-bins ${whisper_mel_bins} --whisper-fbank true
fi
if [ $stage -le 121 ] && [ $stop_stage -ge 121 ]; then if [ $stage -le 121 ] && [ $stop_stage -ge 121 ]; then
log "Stage 121: Prepare MagicData, Primewords, ST-CMDS, THCHS-30 for whisper" log "Stage 121: Prepare MagicData, Primewords, ST-CMDS, THCHS-30 for whisper"