This commit is contained in:
yfyeung 2024-04-02 07:52:11 +00:00
parent f5b8d5dad2
commit 4ae9a00ec5
2 changed files with 3 additions and 3 deletions

View File

@ -17,8 +17,8 @@
import logging
import re
from pathlib import Path
import unicodedata
from pathlib import Path
from lhotse import CutSet, SupervisionSegment
from lhotse.recipes.utils import read_manifests_if_cached
@ -73,7 +73,7 @@ def normalize_text(
text = re.sub(r"\u00B0\u0043", "องศาเซลเซียส", text) # °C
text = re.sub(r"\u00B0\u0046", "องศาฟาเรนไฮต์", text) # °F
text = re.sub(r"\u00B0", "องศา", text) # °
# Remove blank symbols
text = re.sub(r"\s", "", text)

View File

@ -48,7 +48,7 @@ if [ $stage -le 2 ] && [ $stop_stage -ge 2 ]; then
log "Stage 2: Compute fbank for gigaspeech2"
mkdir -p data/fbank
if [ ! -e data/fbank/.gigaspeech2.done ]; then
./local/compute_fbank_gigaspeech2.py
./local/compute_fbank_gigaspeech2.py --lang $lang
touch data/fbank/.gigaspeech2.done
fi
fi