|
@@ -21,7 +21,7 @@ feats_type=fbank
|
|
|
token_type=char
|
|
token_type=char
|
|
|
scp=wav.scp
|
|
scp=wav.scp
|
|
|
type=sound
|
|
type=sound
|
|
|
-stage=1
|
|
|
|
|
|
|
+stage=2
|
|
|
stop_stage=3
|
|
stop_stage=3
|
|
|
|
|
|
|
|
# feature configuration
|
|
# feature configuration
|
|
@@ -100,7 +100,7 @@ if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
|
|
|
echo "<blank>" > ${token_list}
|
|
echo "<blank>" > ${token_list}
|
|
|
echo "<s>" >> ${token_list}
|
|
echo "<s>" >> ${token_list}
|
|
|
echo "</s>" >> ${token_list}
|
|
echo "</s>" >> ${token_list}
|
|
|
- utils/text2token.py -s 1 -n 1 --space "" ${feats_dir}/data/train/text | cut -f 2- -d" " | tr " " "\n" \
|
|
|
|
|
|
|
+ utils/text2token.py -s 1 -n 1 --space "" ${feats_dir}/data/$train_set/text | cut -f 2- -d" " | tr " " "\n" \
|
|
|
| sort | uniq | grep -a -v -e '^\s*$' | awk '{print $0}' >> ${token_list}
|
|
| sort | uniq | grep -a -v -e '^\s*$' | awk '{print $0}' >> ${token_list}
|
|
|
echo "<unk>" >> ${token_list}
|
|
echo "<unk>" >> ${token_list}
|
|
|
fi
|
|
fi
|
嘉渊