diff --git a/code/tokenizer.py b/code/tokenizer.py
new file mode 100644
index 0000000..34d5b16
--- /dev/null
+++ b/code/tokenizer.py
@@ -0,0 +1,64 @@
+#! pip install tokenizers
+
+from tokenizers import Tokenizer
+from tokenizers.models import BPE
+from tokenizers.trainers import BpeTrainer
+from tokenizers.pre_tokenizers import Whitespace
+from tokenizers.processors import BertProcessing
+from tokenizers import normalizers
+from tokenizers.normalizers import NFD, StripAccents, Lowercase
+from tokenizers import pre_tokenizers
+from tokenizers.pre_tokenizers import Digits
+
+# Load data from files
+#paths = [str(x) for x in Path("F:/dockerfile-model/dockerfile_data_1/").glob("*.txt")]
+import pymysql
+#建立数据库连接
+
+def getSQLData(table_name):
+    con = pymysql.connect(host='localhost', user='root', passwd='1234',db='xxx')
+    cur = con.cursor()
+    #用cursor中的execute 使用DDL语句创建一个名为 STUDENT 的表,指定表的字段以及字段类型
+    cur.execute("SELECT content_valid from %s;"%table_name)
+    rows = cur.fetchall()
+    return rows
+
+text = []
+table_names = ['dockerfile']
+for table in table_names:
+    rows = getSQLData(table)
+    for row in rows:
+        #print(row[0].replace('\n',' ').strip())
+        text.append(row[0].replace('\n',' ').strip())
+
+# Initialize a tokenizer
+normalizer = normalizers.Sequence([
+    NFD(),          # Unicode正规化
+    StripAccents()] # remove accents
+)
+
+tokenizer = Tokenizer(BPE(unk_token="<unk>"))
+
+pre_tokenizer = pre_tokenizers.Sequence([
+    Whitespace(),
+    Digits(individual_digits=False)
+])
+
+tokenizer.pre_tokenizer = pre_tokenizer
+
+tokenizer.normalizer = normalizer  # 更新到tokenizer里
+trainer = BpeTrainer(special_tokens=["<s>", "<pad>", "</s>", "<unk>", "<mask>"],vocab_size=50000, show_progress=True, min_frequency=2)
+# Customize training
+tokenizer.train_from_iterator(text, trainer)
+
+#tokenizer.post_processor = BertProcessing(
+#    ("</s>", tokenizer.token_to_id("</s>")),
+#    ("<s>", tokenizer.token_to_id("<s>")),
+#)
+
+#tokenizer.enable_padding(pad_id=1, pad_token="<pad>")
+
+tokenizer.save_model("tokenizer")
+
+print(tokenizer.get_vocab_size())
+