Data Leakage problem

[Khey17]

Hey Everyone,
So I was dealing with a Twitter dataset to build the NLP model and I've split the data as follows:
from sklearn.model_selection import train_test_split

train_sen, val_sen, train_labels, val_labels = train_test_split(train_df['text'].to_numpy(), train_df['target'].to_numpy(), test_size=0.1, random_state=42)
len(train_sen), len(train_labels), len(val_sen), len(val_labels)

len(list(set(train_sen) & set(val_sen))) # no. of common samples
12

Some common samples belong to both train and validation sets. Doesn't that mean the data is leaking? like is this supposed to happen? I don't know what exactly I am missing here… I've checked it cell by cell but still getting the same output. (In this case, I got an accuracy of 82.16…(train) and 82.62…(val))

Is it normal to have some common samples for better generalization or does it lead to overfitting?