diff --git a/egs/gigaspeech/ASR/conformer_ctc/train.py b/egs/gigaspeech/ASR/conformer_ctc/train.py
index f890913db..2965cde18 100755
--- a/egs/gigaspeech/ASR/conformer_ctc/train.py
+++ b/egs/gigaspeech/ASR/conformer_ctc/train.py
@@ -107,7 +107,7 @@ def get_parser():
     parser.add_argument(
         "--lang-dir",
         type=str,
-        default="data/lang_bpe_5000",
+        default="data/lang_bpe_500",
         help="""The lang dir
         It contains language related input files such as
         "lexicon.txt"
diff --git a/egs/gigaspeech/ASR/prepare.sh b/egs/gigaspeech/ASR/prepare.sh
index 559961507..10afe8379 100755
--- a/egs/gigaspeech/ASR/prepare.sh
+++ b/egs/gigaspeech/ASR/prepare.sh
@@ -42,7 +42,6 @@ dl_dir=$PWD/download
 # It will generate data/lang_bpe_xxx,
 # data/lang_bpe_yyy if the array contains xxx, yyy
 vocab_sizes=(
-  5000
   500
 )