data = pd.read_csv('sentiment_data.csv')

train_data, test_data = train_test_split(data, test_size=0.2, random_state=42)

tokenizer = Tokenizer(model_name)
X_train = tokenizer.tokenize(train_data['text'].tolist())
X_test = tokenizer.tokenize(test_data['text'].tolist())

y_train = train_data['sentiment'].map({'negative': 0, 'positive': 1}).values
y_test = test_data['sentiment'].map({'negative': 0, 'positive': 1}).values

        __
        __