update repo

嘉渊

2023-05-16 12dd848db2cfd0e2ae6f32cfb1a5aecdf0f77365

 egs/librispeech_100h/conformer/run.sh

@@ -19,6 +19,7 @@
token_type=bpe
type=sound
scp=wav.scp
speed_perturb="0.9 1.0 1.1"
stage=3
stop_stage=4

@@ -82,6 +83,11 @@
    for x in dev-clean dev-other test-clean test-other train-clean-100; do
        local/data_prep.sh ${raw_data}/LibriSpeech/${x} ${feats_dir}/data/${x//-/_}
    done
    mkdir $feats_dir/data/$valid_set
    dev_sets="dev_clean dev_other"
    for file in wav.scp text; do
        ( for f in $dev_sets; do cat $feats_dir/data/$f/$file; done ) | sort -k1 > $feats_dir/data/$valid_set/$file || exit 1;
    done
fi

if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
@@ -133,10 +139,11 @@
                --data_dir ${feats_dir}/data \
                --train_set ${train_set} \
                --valid_set ${valid_set} \
                --cmvn_file ${feats_dir}/data/${train_set}/cmvn/cmvn.mvn \
                --speed_perturb ${speed_perturb} \
                --resume true \
                --output_dir ${exp_dir}/exp/${model_dir} \
                --config $asr_config \
                --input_size $feats_dim \
                --ngpu $gpu_num \
                --num_worker_count $count \
                --multiprocessing_distributed true \
@@ -162,7 +169,7 @@
            exit 0
        fi
        mkdir -p "${_logdir}"
        _data="${feats_dir}/${dumpdir}/${dset}"
        _data="${feats_dir}/data/${dset}"
        key_file=${_data}/${scp}
        num_scp_file="$(<${key_file} wc -l)"
        _nj=$([ $inference_nj -le $num_scp_file ] && echo "$inference_nj" || echo "$num_scp_file")