diff --git a/egs/gigaspeech/ASR/local/compute_fbank_gigaspeech.py b/egs/gigaspeech/ASR/local/compute_fbank_gigaspeech.py new file mode 100755 index 000000000..ea5d3dc6e --- /dev/null +++ b/egs/gigaspeech/ASR/local/compute_fbank_gigaspeech.py @@ -0,0 +1,99 @@ +#!/usr/bin/env python3 +# Copyright 2021 Xiaomi Corp. (authors: Fangjun Kuang) +# +# See ../../../../LICENSE for clarification regarding multiple authors +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + + +""" +This file computes fbank features of the GigaSpeech dataset. +It looks for manifests in the directory data/manifests. + +The generated fbank features are saved in data/fbank. +""" + +import logging +import os +from pathlib import Path + +import torch +from lhotse import CutSet, Fbank, FbankConfig, LilcomHdf5Writer +from lhotse.recipes.utils import read_manifests_if_cached + +from icefall.utils import get_executor + +# Torch's multithreaded behavior needs to be disabled or +# it wastes a lot of CPU and slow things down. +# Do this outside of main() in case it needs to take effect +# even when we are not invoking the main (e.g. when spawning subprocesses). +torch.set_num_threads(1) +torch.set_num_interop_threads(1) + + +def compute_fbank_gigaspeech(): + src_dir = Path("data/manifests") + output_dir = Path("data/fbank") + num_jobs = min(10, os.cpu_count()) + num_mel_bins = 80 + + dataset_parts = ( + "XL", + "DEV", + "TEST", + ) + manifests = read_manifests_if_cached( + dataset_parts=dataset_parts, + output_dir=src_dir, + prefix="gigaspeech", + suffix="jsonl.gz", + ) + assert manifests is not None + + extractor = Fbank(FbankConfig(num_mel_bins=num_mel_bins)) + + with get_executor() as ex: # Initialize the executor only once. + for partition, m in manifests.items(): + if (output_dir / f"cuts_{partition}.jsonl.gz").is_file(): + logging.info(f"{partition} already exists - skipping.") + continue + logging.info(f"Processing {partition}") + cut_set = CutSet.from_manifests( + recordings=m["recordings"], + supervisions=m["supervisions"], + ) + if "train" in partition: + cut_set = ( + cut_set + + cut_set.perturb_speed(0.9) + + cut_set.perturb_speed(1.1) + ) + cut_set = cut_set.compute_and_store_features( + extractor=extractor, + storage_path=f"{output_dir}/feats_{partition}", + # when an executor is specified, make more partitions + num_jobs=num_jobs if ex is None else 80, + executor=ex, + storage_type=LilcomHdf5Writer, + ) + cut_set.to_json(output_dir / f"cuts_{partition}.jsonl.gz") + + +if __name__ == "__main__": + formatter = ( + "%(asctime)s %(levelname)s [%(filename)s:%(lineno)d] %(message)s" + ) + + logging.basicConfig(format=formatter, level=logging.INFO) + + compute_fbank_gigaspeech() diff --git a/egs/gigaspeech/ASR/local/compute_fbank_musan.py b/egs/gigaspeech/ASR/local/compute_fbank_musan.py new file mode 120000 index 000000000..5833f2484 --- /dev/null +++ b/egs/gigaspeech/ASR/local/compute_fbank_musan.py @@ -0,0 +1 @@ +../../../librispeech/ASR/local/compute_fbank_musan.py \ No newline at end of file diff --git a/egs/gigaspeech/ASR/prepare.sh b/egs/gigaspeech/ASR/prepare.sh index e69de29bb..fd6fb217f 100755 --- a/egs/gigaspeech/ASR/prepare.sh +++ b/egs/gigaspeech/ASR/prepare.sh @@ -0,0 +1,234 @@ +#!/usr/bin/env bash + +set -eou pipefail + +nj=15 +stage=0 +stop_stage=100 + +# We assume dl_dir (download dir) contains the following +# directories and files. If not, they will be downloaded +# by this script automatically. +# +# - $dl_dir/GigaSpeech +# You can find audio, dict, GigaSpeech.json inside it. +# You can apply for the download credentials by following +# https://github.com/SpeechColab/GigaSpeech#dataset-download +# +# - $dl_dir/musan +# This directory contains the following directories downloaded from +# http://www.openslr.org/17/ +# +# - music +# - noise +# - speech +dl_dir=$PWD/download + +. 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 "dl_dir: $dl_dir" + +if [ $stage -le 0 ] && [ $stop_stage -ge 0 ]; then + log "Stage 0: Download data" + + [ ! -e $dl_dir/GigaSpeech ] && mkdir -p $dl_dir/GigaSpeech + + # If you have pre-downloaded it to /path/to/GigaSpeech, + # you can create a symlink + # + # ln -sfv /path/to/GigaSpeech $dl_dir/GigaSpeech + # + if [ ! -d $dl_dir/GigaSpeech/audio ] && [ ! -f $dl_dir/GigaSpeech.json ]; then + # Check credentials. + if [ ! -f $dl_dir/password ]; then + echo -n "$0: Please apply for the download credentials by following" + echo -n "https://github.com/SpeechColab/GigaSpeech#dataset-download" + echo " and save it to $dl_dir/password." + exit 1; + fi + PASSWORD=`cat $dl_dir/password 2>/dev/null` + if [ -z "$PASSWORD" ]; then + echo "$0: Error, $dl_dir/password is empty." + exit 1; + fi + PASSWORD_MD5=`echo $PASSWORD | md5sum | cut -d ' ' -f 1` + if [[ $PASSWORD_MD5 != "dfbf0cde1a3ce23749d8d81e492741b8" ]]; then + echo "$0: Error, invalid $dl_dir/password." + exit 1; + fi + lhotse download gigaspeech --subset auto --host tsinghua \ + $dl_dir/password $dl_dir/GigaSpeech + fi + + # If you have pre-downloaded it to /path/to/musan, + # you can create a symlink + # + # ln -sfv /path/to/musan $dl_dir/ + # + if [ ! -d $dl_dir/musan ]; then + lhotse download musan $dl_dir + fi +fi + +if [ $stage -le 1 ] && [ $stop_stage -ge 1 ]; then + log "Stage 1: Prepare GigaSpeech manifest" + # We assume that you have downloaded the GigaSpeech corpus + # to $dl_dir/GigaSpeech + mkdir -p data/manifests + lhotse prepare gigaspeech --subset auto -j $nj \ + $dl_dir/GigaSpeech data/manifests +fi + +if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then + log "Stage 2: Prepare musan manifest" + # We assume that you have downloaded the musan corpus + # to $dl_dir/musan + mkdir -p data/manifests + lhotse prepare musan $dl_dir/musan data/manifests +fi + +if [ $stage -le 3 ] && [ $stop_stage -ge 3 ]; then + log "Stage 3: Compute fbank for GigaSpeech" + mkdir -p data/fbank + ./local/compute_fbank_gigaspeech.py +fi + +if [ $stage -le 4 ] && [ $stop_stage -ge 4 ]; then + log "Stage 4: Compute fbank for musan" + mkdir -p data/fbank + ./local/compute_fbank_musan.py +fi + +if [ $stage -le 5 ] && [ $stop_stage -ge 5 ]; then + log "Stage 5: Prepare phone based lang" + lang_dir=data/lang_phone + mkdir -p $lang_dir + + (echo '!SIL SIL'; echo ' SPN'; echo ' SPN'; ) | + cat - $dl_dir/lm/librispeech-lexicon.txt | + sort | uniq > $lang_dir/lexicon.txt + + if [ ! -f $lang_dir/L_disambig.pt ]; then + ./local/prepare_lang.py --lang-dir $lang_dir + fi +fi + + +if [ $stage -le 6 ] && [ $stop_stage -ge 6 ]; then + log "Stage 6: Prepare BPE based lang" + + for vocab_size in ${vocab_sizes[@]}; do + lang_dir=data/lang_bpe_${vocab_size} + mkdir -p $lang_dir + # We reuse words.txt from phone based lexicon + # so that the two can share G.pt later. + cp data/lang_phone/words.txt $lang_dir + + if [ ! -f $lang_dir/transcript_words.txt ]; then + log "Generate data for BPE training" + files=$( + find "$dl_dir/LibriSpeech/train-clean-100" -name "*.trans.txt" + find "$dl_dir/LibriSpeech/train-clean-360" -name "*.trans.txt" + find "$dl_dir/LibriSpeech/train-other-500" -name "*.trans.txt" + ) + for f in ${files[@]}; do + cat $f | cut -d " " -f 2- + done > $lang_dir/transcript_words.txt + fi + + ./local/train_bpe_model.py \ + --lang-dir $lang_dir \ + --vocab-size $vocab_size \ + --transcript $lang_dir/transcript_words.txt + + if [ ! -f $lang_dir/L_disambig.pt ]; then + ./local/prepare_lang_bpe.py --lang-dir $lang_dir + fi + done +fi + +if [ $stage -le 7 ] && [ $stop_stage -ge 7 ]; then + log "Stage 7: Prepare bigram P" + + for vocab_size in ${vocab_sizes[@]}; do + lang_dir=data/lang_bpe_${vocab_size} + + if [ ! -f $lang_dir/transcript_tokens.txt ]; then + ./local/convert_transcript_words_to_tokens.py \ + --lexicon $lang_dir/lexicon.txt \ + --transcript $lang_dir/transcript_words.txt \ + --oov "" \ + > $lang_dir/transcript_tokens.txt + fi + + if [ ! -f $lang_dir/P.arpa ]; then + ./shared/make_kn_lm.py \ + -ngram-order 2 \ + -text $lang_dir/transcript_tokens.txt \ + -lm $lang_dir/P.arpa + fi + + if [ ! -f $lang_dir/P.fst.txt ]; then + python3 -m kaldilm \ + --read-symbol-table="$lang_dir/tokens.txt" \ + --disambig-symbol='#0' \ + --max-order=2 \ + $lang_dir/P.arpa > $lang_dir/P.fst.txt + fi + done +fi + +if [ $stage -le 8 ] && [ $stop_stage -ge 8 ]; then + log "Stage 8: Prepare G" + # We assume you have install kaldilm, if not, please install + # it using: pip install kaldilm + + mkdir -p data/lm + if [ ! -f data/lm/G_3_gram.fst.txt ]; then + # It is used in building HLG + python3 -m kaldilm \ + --read-symbol-table="data/lang_phone/words.txt" \ + --disambig-symbol='#0' \ + --max-order=3 \ + $dl_dir/lm/3-gram.pruned.1e-7.arpa > data/lm/G_3_gram.fst.txt + fi + + if [ ! -f data/lm/G_4_gram.fst.txt ]; then + # It is used for LM rescoring + python3 -m kaldilm \ + --read-symbol-table="data/lang_phone/words.txt" \ + --disambig-symbol='#0' \ + --max-order=4 \ + $dl_dir/lm/4-gram.arpa > data/lm/G_4_gram.fst.txt + fi +fi + +if [ $stage -le 9 ] && [ $stop_stage -ge 9 ]; then + log "Stage 9: Compile HLG" + ./local/compile_hlg.py --lang-dir data/lang_phone + + for vocab_size in ${vocab_sizes[@]}; do + lang_dir=data/lang_bpe_${vocab_size} + ./local/compile_hlg.py --lang-dir $lang_dir + done +fi