[Solved] Unlock Python Keras Error: AttributeError Guide in 2024: ‘Sequential‘ object has no attribute ‘predict_classes'

[Solved] Unlock Python Keras Error: AttributeError Guide in 2024: ‘Sequential‘ object has no attribute ‘predict_classes’

Author: Amresh Mishra | Published On: February 25, 2024

This article is about Python Keras Error: AttributeError using Keras in Python to execute yhat_classes = model.predict_classes(X_test) code with an error: AttributeError: ‘Sequential’ object has no attribute ‘predict_classes’ solution.

model = Sequential()
model.add(Dense(24, input_dim=13, activation='relu'))
model.add(Dense(18, activation='relu'))
model.add(Dense(6, activation='softmax'))
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
history = model.fit(X_train, y_train, batch_size = 256, epochs = 10, verbose = 2, validation_split = 0.2)
score, acc = model.evaluate(X_test, y_test,verbose=2, batch_size= 256)
print('test accuracy:', acc)
yhat_classes = model.predict_classes(X_test)
Python Keras Error: AttributeError: ‘Sequential‘ object has no attribute ‘predict_classes'

This prediction was removed in the TensorFlow version 2.6_ Classes function.

Reference documents: https://keras.rstudio.com/reference/predict_proba.html#details

The following codes can be used:



You can also try tensorflow 2.5 or other versions to solve this problem.

Using tensorflow version 2.5, there may be the following warning messages:

tensorflow\python\keras\engine\sequential.py:455: UserWarning: model.predict_classes() is deprecated and will be removed after 2021-01-01. Please use instead:* np.argmax(model.predict(x), axis=-1), if your model does multi-class classification (e.g. if it uses a softmax last-layer activation).* (model.predict(x) > 0.5).astype("int32"), if your model does binary classification (e.g. if it uses a sigmoid last-layer activation).

Introduction to Python Keras Error: AttributeError

Python Keras Error: AttributeError: Welcome, intrepid coder, to the labyrinth of Python and Keras, where the path to machine learning glory is paved with perplexing error messages. Today’s quest? Conquering the formidable AttributeError: ‘Sequential‘ object has no attribute ‘predict_classes’. But fret not, brave soul, for we shall embark on this journey together, armed with wit, wisdom, and an unyielding spirit!

Unraveling the Enigma: Python Keras Error: AttributeError
Let us first decipher the cryptic message that has halted your coding crusade. In the realm of Keras, the Sequential model API reigns supreme, allowing you to craft intricate neural networks with ease. Yet, when faced with the dreaded AttributeError, Keras seems to scoff at your attempts to wield the ‘predict_classes’ attribute, as if it were a forbidden incantation lost to the sands of time.

The Quandary Unveiled:
You may find yourself pondering, “But wait, haven’t I seen countless tutorials and examples where ‘predict_classes’ was used without a hitch?” Ah, my fellow traveler, therein lies the twist in our tale. Keras, much like a capricious deity, evolves with the tides of time. What once worked flawlessly may now elude us, hidden behind the veil of updates and version discrepancies.

Charting the Course to Resolution:
Now, dear comrade, it is time to arm ourselves with knowledge and devise a strategy to vanquish this error once and for all. Fear not, for I shall impart upon you the sacred wisdom passed down through generations of intrepid coders. Instead of futilely grasping at the elusive ‘predict_classes’, let us embrace a different approach, akin to navigating a labyrinth with a map forged of pure logic.

Behold the Solution:
Cast aside your doubts and behold the elegance of the solution that awaits us. By harnessing the power of the ‘predict’ method bestowed upon us by Keras, and wielding the mystical arts of numpy, we shall discern the class with the highest probability, as a sage discerns truth from falsehood amidst a sea of uncertainty.

import numpy as np

# Assume 'model' is your trained Keras Sequential model
predictions = model.predict(test_data)
predicted_classes = np.argmax(predictions, axis=1)

With but a few lines of code, we defy the odds and emerge triumphant, our victory proclaimed to the heavens.

Also Read:

Frequently Asked Inquiries:

Q: Why does Keras insist on confounding us with ever-changing quirks?

A: Ah, the eternal question that plagues both novice and seasoned coder alike! Keras, in its infinite wisdom, dances to the beat of its own drum, forever evolving in pursuit of perfection. Embrace the chaos, dear friend, for therein lies the thrill of the coding odyssey.

Q: Can I not simply cling to the comfort of ‘predict_classes’ and be done with it?

A: Tempting though it may be, to cling to the familiar is to deny oneself the opportunity for growth and enlightenment. Embrace the unknown, for therein lies the path to true mastery.

Q: Is there a panacea to shield us from the sting of such errors in the future?

A: Alas, my friend, the road to coding enlightenment is paved with trials and tribulations. Remain vigilant, consult the sacred texts (i.e., documentation), and trust in the collective wisdom of the coding community.


As we reach the culmination of our quest, let us reflect upon the journey that has brought us to this moment. Through perseverance, ingenuity, and perhaps a touch of madness, we have transcended the shackles of the AttributeError and emerged as champions of the code. Fear not the errors that lie ahead, for armed with knowledge and camaraderie, we shall face them head-on, undaunted by the challenges that await.

Author: Amresh Mishra
Amresh Mishra is a passionate coder and technology enthusiast dedicated to exploring the vast world of programming. With a keen interest in web development, software engineering, and emerging technologies, Amresh is on a mission to share his knowledge and experience with fellow enthusiasts through his website, CodersCanteen.com.

Leave a Comment