Regression error with Python tensorflow keras

Question

could someone please help me to fix the following error :

[AttributeError: 'super' object has no attribute 'sklearn_tags']

based on my code :

from tensorflow import keras 
from scikeras.wrappers import KerasRegressor
def build_model():
    model = keras.Sequential([
        keras.layers.Dense(32, activation = 'tanh', input_shape = [len(X.keys())]),
        keras.layers.Dense(16, activation = 'tanh'),
        keras.layers.Dense(1, activation = "sigmoid")])
    optimizer = keras.optimizers.RMSprop()
    model.compile(loss = keras.losses.BinaryCrossentropy(),
                  optimizer = optimizer,
                  metrics = [keras.metrics.BinaryAccuracy()])
    return model
kr = KerasRegressor(build_model, epochs=100, batch_size=10000,verbose=1)
kr.fit(X_train,y_train, validation_split = 0.2)

I'm looking everywhere to find a solution but nothing works until now ... Thanks

Answer 1

There is an sklearn issue. In sklearn 1.6.1 the error was turned into warning. You can install sklearn >=1.6.1,<1.7 and just expect DeprecationWarning regarding this issue.

Or another way, you can downgrade to 1.3.1 to avoid this issue

!pip uninstall -y scikit-learn
!pip install scikit-learn==1.3.1