run_tedlium3_lm.sh

#!/bin/bash

# Copyright 2019 Nagoya University (Masao Someki)
# Copyright 2019 Shigeki Karita
#  Apache 2.0  (http://www.apache.org/licenses/LICENSE-2.0)

. ./path.sh || exit 1;
. ./cmd.sh || exit 1;

# general configuration
backend=pytorch
stage=-1       # start from -1 if you need to start from data download
stop_stage=100
ngpu=0         # number of gpus ("0" uses cpu, otherwise use gpu)
debugmode=1
dumpdir=dump   # directory to dump full features
N=0            # number of minibatches to be used (mainly for debugging). "0" uses all minibatches.
verbose=0      # verbose option
resume=        # Resume the training from snapshot

# feature configuration
do_delta=false

# rnnlm related
lm_resume=        # specify a snapshot file to resume LM training
lmtag=            # tag for managing LMs

# decoding parameter
recog_model=model.acc.best # set a model to be used for decoding: 'model.acc.best' or 'model.loss.best'
n_average=10

# bpemode (unigram or bpe)
nbpe=500
bpemode=unigram

# exp tag
tag="" # tag for managing experiments.

train_config=conf/train.yaml
decode_config=conf/decode.yaml
preprocess_config=conf/specaug.yaml
lm_config=conf/lm.yaml

. utils/parse_options.sh || exit 1;

# Set bash to 'debug' mode, it will exit on :
# -e 'error', -u 'undefined variable', -o ... 'error in pipeline', -x 'print commands',
set -e
set -u
set -o pipefail


# legacy setup
data_type=legacy
train_set=train_trim
train_dev=dev_trim
recog_set="dev test"

#adaptation setup
#data_type=speaker-adaptation
#train_set=train_adapt_trim_sp
#train_dev=dev_adapt_trim
#recog_set="dev_adapt test_adapt"


if [ ${stage} -le -1 ] && [ ${stop_stage} -ge -1 ]; then
    echo "stage -1: Data Download"
    local/download_data.sh
fi

if [ ${stage} -le 0 ] && [ ${stop_stage} -ge 0 ]; then
    ### Task dependent. You have to make data the following preparation part by yourself.
    ### But you can utilize Kaldi recipes in most cases
    echo "stage 0: Data preparation"
    local/prepare_data.sh $data_type
    for dset in dev test train; do
    utils/data/modify_speaker_info.sh --seconds-per-spk-max 180 data/${dset}.orig data/${dset}
    done
fi

feat_tr_dir=${dumpdir}/${train_set}/delta${do_delta}; mkdir -p ${feat_tr_dir}
feat_dt_dir=${dumpdir}/${train_dev}/delta${do_delta}; mkdir -p ${feat_dt_dir}
if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
    ### Task dependent. You have to design training and dev sets by yourself.
    ### But you can utilize Kaldi recipes in most cases
    echo "stage 1: Feature Generation"
    fbankdir=fbank
    # # Generate the fbank features; by default 80-dimensional fbanks with pitch on each frame
    # for x in test dev train; do
    #     steps/make_fbank_pitch.sh --cmd "$train_cmd" --nj 64 --write_utt2num_frames true \
    #         data/${x} exp/make_fbank/${x} ${fbankdir}
    #     utils/fix_data_dir.sh data/${x}
    # done

    # remove utt having > 2000 frames or < 10 frames or
    # remove utt having > 400 characters or 0 characters
    remove_longshortdata.sh --maxchars 400 data/train data/train_trim
    remove_longshortdata.sh --maxchars 400 data/dev data/${train_dev}

    # # speed-perturbed
    # utils/perturb_data_dir_speed.sh 0.9 data/train_trim data/temp1
    # utils/perturb_data_dir_speed.sh 1.0 data/train_trim data/temp2
    # utils/perturb_data_dir_speed.sh 1.1 data/train_trim data/temp3
    # utils/combine_data.sh --extra-files utt2uniq data/${train_set} data/temp1 data/temp2 data/temp3
    # rm -r data/temp1 data/temp2 data/temp3
    # steps/make_fbank_pitch.sh --cmd "$train_cmd" --nj 64 --write_utt2num_frames true \
    #     data/${train_set} exp/make_fbank/${train_set} ${fbankdir}
    # utils/fix_data_dir.sh data/${train_set}

    # # compute global CMVN
    # compute-cmvn-stats scp:data/${train_set}/feats.scp data/${train_set}/cmvn.ark

    # # dump features for training
    # dump.sh --cmd "$train_cmd" --nj 32 --do_delta ${do_delta} \
    #     data/${train_set}/feats.scp data/${train_set}/cmvn.ark exp/dump_feats/train ${feat_tr_dir}
    # dump.sh --cmd "$train_cmd" --nj 32 --do_delta ${do_delta} \
    #     data/${train_dev}/feats.scp data/${train_set}/cmvn.ark exp/dump_feats/dev ${feat_dt_dir}
    # for rtask in ${recog_set}; do
    #     feat_recog_dir=${dumpdir}/${rtask}/delta${do_delta}; mkdir -p ${feat_recog_dir}
    #     dump.sh --cmd "$train_cmd" --nj 32 --do_delta ${do_delta} \
    #         data/${rtask}/feats.scp data/${train_set}/cmvn.ark exp/dump_feats/recog/${rtask} \
    #         ${feat_recog_dir}
    # done
fi

dict=data/lang_char/${train_set}_${bpemode}${nbpe}_units.txt
bpemodel=data/lang_char/${train_set}_${bpemode}${nbpe}
echo "dictionary: ${dict}"
if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
    ### Task dependent. You have to check non-linguistic symbols used in the corpus.
    echo "stage 2: Dictionary and Json Data Preparation"
    mkdir -p data/lang_char/
    echo "<unk> 1" > ${dict} # <unk> must be 1, 0 will be used for "blank" in CTC
    cut -f 2- -d" " data/${train_set}/text > data/lang_char/input.txt
    spm_train --input=data/lang_char/input.txt --vocab_size=${nbpe} --model_type=${bpemode} \
        --model_prefix=${bpemodel%%.*} --input_sentence_size=100000000
    spm_encode --model=${bpemodel}.model --output_format=piece < data/lang_char/input.txt | \
        tr ' ' '\n' | sort | uniq | awk '{print $0 " " NR+1}' >> ${dict}
    wc -l ${dict}

    # # make json labels
    # data2json.sh --feat ${feat_tr_dir}/feats.scp --bpecode ${bpemodel}.model \
    #      data/${train_set} ${dict} > ${feat_tr_dir}/data_${bpemode}${nbpe}.json
    # data2json.sh --feat ${feat_dt_dir}/feats.scp --bpecode ${bpemodel}.model \
    #      data/${train_dev} ${dict} > ${feat_dt_dir}/data_${bpemode}${nbpe}.json
    # for rtask in ${recog_set}; do
    #     feat_recog_dir=${dumpdir}/${rtask}/delta${do_delta}
    #     data2json.sh --feat ${feat_recog_dir}/feats.scp --bpecode ${bpemodel}.model \
    #         data/${rtask} ${dict} > ${feat_recog_dir}/data_${bpemode}${nbpe}.json
    # done
fi

# It takes a few days. If you just want to end-to-end ASR without LM,
# you can skip this and remove --rnnlm option in the recognition (stage 5)
if [ -z ${lmtag} ]; then
    lmtag=$(basename ${lm_config%.*})
fi
lmexpname=train_rnnlm_${backend}_${lmtag}_${bpemode}${nbpe}
lmexpdir=exp/${lmexpname}
mkdir -p ${lmexpdir}

if [ ${stage} -le 3 ] && [ ${stop_stage} -ge 3 ]; then
    echo "stage 3: LM Preparation"
    lmdatadir=data/local/lm_train_${bpemode}${nbpe}
    if [ ! -e ${lmdatadir} ]; then
        mkdir -p ${lmdatadir}
        gunzip -c db/TEDLIUM_release-3/LM/*.en.gz | sed 's/ <\/s>//g' | local/join_suffix.py |\
            spm_encode --model=${bpemodel}.model --output_format=piece > ${lmdatadir}/train.txt
        cut -f 2- -d" " data/${train_dev}/text | spm_encode --model=${bpemodel}.model --output_format=piece \
                                                            > ${lmdatadir}/valid.txt
    fi

    ${cuda_cmd} --gpu ${ngpu} ${lmexpdir}/train.log \
        lm_train.py \
        --config ${lm_config} \
        --ngpu ${ngpu} \
        --backend ${backend} \
        --verbose 1 \
        --outdir ${lmexpdir} \
        --tensorboard-dir tensorboard/${lmexpname} \
        --train-label ${lmdatadir}/train.txt \
        --valid-label ${lmdatadir}/valid.txt \
        --resume ${lm_resume} \
        --dict ${dict} \
        --dump-hdf5-path ${lmdatadir}
fi

if [ -z ${tag} ]; then
    expname=${train_set}_${backend}_nbpe${nbpe}_ngpu${ngpu}_$(basename ${train_config%.*})
    if ${do_delta}; then
        expname=${expname}_delta
    fi
    if [ -n "${preprocess_config}" ]; then
	expname=${expname}_$(basename ${preprocess_config%.*})
    fi
else
    expname=${train_set}_${backend}_${tag}
fi
expdir=exp/${expname}
mkdir -p ${expdir}

if [ ${stage} -le 4 ] && [ ${stop_stage} -ge 4 ]; then
    echo "stage 4: Network Training"
    ${cuda_cmd} --gpu ${ngpu} ${expdir}/train.log \
        asr_train.py \
        --ngpu ${ngpu} \
        --preprocess-conf ${preprocess_config} \
        --config ${train_config} \
        --backend ${backend} \
        --outdir ${expdir}/results \
        --tensorboard-dir tensorboard/${expname} \
        --debugmode ${debugmode} \
        --dict ${dict} \
        --debugdir ${expdir} \
        --minibatches ${N} \
        --verbose ${verbose} \
        --resume ${resume} \
        --train-json ${feat_tr_dir}/data_${bpemode}${nbpe}.json \
        --valid-json ${feat_dt_dir}/data_${bpemode}${nbpe}.json
fi

if [ ${stage} -le 5 ] && [ ${stop_stage} -ge 5 ]; then
    echo "stage 5: Decoding"
    nj=32
    if [[ $(get_yaml.py ${train_config} model-module) = *transformer* ]]; then
	recog_model=model.last${n_average}.avg.best
	average_checkpoints.py --backend ${backend} \
			       --snapshots ${expdir}/results/snapshot.ep.* \
			       --out ${expdir}/results/${recog_model} \
			       --num ${n_average}
    fi
    pids=() # initialize pids
    for rtask in ${recog_set}; do
    (
        decode_dir=decode_${rtask}_$(basename ${decode_config%.*})
        feat_recog_dir=${dumpdir}/${rtask}/delta${do_delta}

        # split data
        splitjson.py --parts ${nj} ${feat_recog_dir}/data_${bpemode}${nbpe}.json

        #### use CPU for decoding
        ngpu=0
        ${decode_cmd} JOB=1:${nj} ${expdir}/${decode_dir}/log/decode.JOB.log \
            asr_recog.py \
            --config ${decode_config} \
            --ngpu ${ngpu} \
            --backend ${backend} \
            --debugmode ${debugmode} \
            --verbose ${verbose} \
            --recog-json ${feat_recog_dir}/split${nj}utt/data_${bpemode}${nbpe}.JOB.json \
            --result-label ${expdir}/${decode_dir}/data.JOB.json \
            --model ${expdir}/results/${recog_model}  \
            --rnnlm ${lmexpdir}/rnnlm.model.best

        score_sclite.sh --bpe ${nbpe} --bpemodel ${bpemodel}.model --wer true ${expdir}/${decode_dir} ${dict}

    ) &
    pids+=($!) # store background pids
    done
    i=0; for pid in "${pids[@]}"; do wait ${pid} || ((++i)); done
    [ ${i} -gt 0 ] && echo "$0: ${i} background jobs are failed." && false
    echo "Finished"
fi