Normalizing data in python in Keras Linear regression Neural Network

Question

I am developing the Neural Network model for linear regression using Keras in Python. I have divided my master dataframe in two parts. (1) Features and (2) Targets

then by using below code, I have made splits of my features and targets into train and test sets.

from sklearn.model_selection import train_test_split
x = df_200[features]
y = df_200[targets]
x_train, x_test, y_train, y_test = train_test_split(x,y, random_state = 1, test_size = 0.2)

But my dataset is having wide ranges of numbers. For which I am thinking to normalize the data to minimize the loss. I would like to ask you that, how would I normalize the data? and Should I normalize the whole dataset ( features and inputs) or only training dataset (x_train and y_tain)??