mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-08-10 02:22:17 +00:00
97 lines
2.7 KiB
Bash
Executable File
97 lines
2.7 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
|
|
# fix segmentation fault reported in https://github.com/k2-fsa/icefall/issues/674
|
|
export PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python
|
|
|
|
set -eou pipefail
|
|
|
|
nj=15
|
|
stage=-1
|
|
stop_stage=100
|
|
|
|
# We assume dl_dir (download dir) contains the following
|
|
# directories and files. Most of them can't be downloaded automatically
|
|
# as they are not publically available and require a license purchased
|
|
# from the LDC.
|
|
#
|
|
# - $dl_dir/musan
|
|
# This directory contains the following directories downloaded from
|
|
# http://www.openslr.org/17/
|
|
#
|
|
# - music
|
|
# - noise
|
|
# - speech
|
|
|
|
dl_dir=./download
|
|
# swbd1_dir="/export/corpora3/LDC/LDC97S62"
|
|
swbd1_dir=./download/LDC97S62/
|
|
|
|
# eval2000_dir contains the following files and directories
|
|
# downloaded from LDC website:
|
|
# - LDC2002S09
|
|
# - hub5e_00
|
|
# - LDC2002T43
|
|
# - reference
|
|
eval2000_dir="/export/corpora2/LDC/eval2000"
|
|
|
|
rt03_dir="/export/corpora/LDC/LDC2007S10"
|
|
fisher_dir="/export/corpora3/LDC/LDC2004T19"
|
|
|
|
. shared/parse_options.sh || exit 1
|
|
|
|
# vocab size for sentence piece models.
|
|
# It will generate data/lang_bpe_xxx,
|
|
# data/lang_bpe_yyy if the array contains xxx, yyy
|
|
vocab_sizes=(
|
|
# 5000
|
|
# 2000
|
|
1000
|
|
500
|
|
)
|
|
|
|
# All files generated by this script are saved in "data".
|
|
# You can safely remove "data" and rerun this script to regenerate it.
|
|
mkdir -p data
|
|
|
|
log() {
|
|
# This function is from espnet
|
|
local fname=${BASH_SOURCE[1]##*/}
|
|
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
|
|
}
|
|
|
|
log "swbd1_dir: $swbd1_dir"
|
|
log "eval2000_dir: $eval2000_dir"
|
|
log "rt03_dir: $rt03_dir"
|
|
|
|
if [ $stage -le 1 ] && [ $stop_stage -ge 1 ]; then
|
|
log "Stage 1 I: Compute narrowband fbank for SwitchBoard"
|
|
if [ ! -e data/fbank_nb/.swbd.done ]; then
|
|
mkdir -p data/fbank_nb/swbd_split${num_splits}/
|
|
for index in $(seq 1 16); do
|
|
./local/compute_fbank_swbd_nb.py --split-index ${index} &
|
|
done
|
|
wait
|
|
pieces=$(find data/fbank_nb/swbd_split${num_splits} -name "swbd_cuts_all.*.jsonl.gz")
|
|
lhotse combine $pieces data/fbank_nb/swbd_cuts_all.jsonl.gz
|
|
touch data/fbank_nb/.swbd.done
|
|
fi
|
|
fi
|
|
|
|
if [ $stage -le 1 ] && [ $stop_stage -ge 1 ]; then
|
|
log "Stage 1 II: Compute narrowband fbank for eval2000"
|
|
if [ ! -e data/fbank_nb/.eval2000.done ]; then
|
|
mkdir -p data/fbank_nb/eval2000/
|
|
./local/compute_fbank_eval2000_nb.py
|
|
touch data/fbank_nb/.eval2000.done
|
|
fi
|
|
fi
|
|
|
|
if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then
|
|
log "Stage 2: Compute narrowband fbank for musan"
|
|
mkdir -p data/fbank_nb/
|
|
if [ ! -e data/fbank_nb/.musan.done ]; then
|
|
./local/compute_fbank_musan_nb.py
|
|
touch data/fbank_nb/.musan.done
|
|
fi
|
|
fi
|