How do I do a validation split on keras dataset

Question

At the moment the code is splitting the dataset in half, 50% for training and 50% for test, how could i split the data in other proportions like 80/20?

(X_train, y_train), (X_test, y_test) = imdb.load_data(num_words=top_words)

i have added the validation_split function in the model.compile section of my CNN model but that only splits the training dataset and not the whole data set.

model.fit(X_train, y_train,validation_split=0.2, epochs=3, callbacks=[
          tensorBoardCallback], batch_size=64)

i have found that the sklearn train_test_split() is the common way to do it but when i put the code from the website in my code i get the error:

----> 9 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

NameError: name 'X' is not defined 

the code responsible:

# Using keras to load the dataset with the top_words
top_words = 10000
(X_train, y_train), (X_test, y_test) = imdb.load_data(num_words=top_words)
word_index = keras.datasets.imdb.get_word_index()

# Split into training and testing data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

Answer 1

# Combine the data and labels and then do the split.

(X_train, y_train), (X_test, y_test) = imdb.load_data(num_words=top_words)

X = np.concatenate((X_train, X_test), axis=0)

y = np.concatenate((y_train, y_test), axis=0)

# Split into training and testing data

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)