The "show_accuracy" argument is deprecated

[jdelange]

C:\Program Files\Anaconda2\lib\site-packages\keras\models.py:635: UserWarning:
he "show_accuracy" argument is deprecated, instead you should pass the "accurac
" metric to the model at compile time:
model.compile(optimizer, loss, metrics=["accuracy"])
warnings.warn('The "show_accuracy" argument is deprecated, '

In your example application:

model.fit(X_train, Y_train,
              batch_size={{choice([64, 128])}},
              nb_epoch=1,
              show_accuracy=True,
              verbose=2,
              validation_data=(X_test, Y_test))

[isrugeek]

@jdelange How did you fix this issue i also got this Issue in my Project ( Not this Code)
Thanks in Advance

[jdelange]

@isrugeek As noted above: replace the show_accuracy=True statement with metrics=['accuracy'] I'm not sure exactly where this applies, but it is either in the code you are writing, or it is in existing (example) code in Keras/Hyperas if it has not yet been fixed. Note: I had too many issues with Hyperas, I used Hyperopt instead.

[isrugeek]

@jdelange Thanks but after i changed it shows me another error

TypeError: Received unknown keyword arguments: {'metrics': ['accuracy']}

[jdelange]

@isrugeek Sorry, can't see what you are doing. The syntax provided is correct, so the call may not be applicable/available in your case? Sounds quite basic. Not sure how experienced you are. Have a look at this site Machine learning mastery lots of basic examples to get started. Helped me a lot.

[isrugeek]

@jdelange Thanks buddy. i am confused because it works on my different machine and not on MacBook . any how i fixed the problem by edit the Model.py( Keras File) by comparing my Old machine.

Thanks

[JAStark]

The issue arises from:
model.fit(X_train, y_train_ohe, nb_epoch=50, batch_size=300, verbose=1, validation_split=0.1, show_metrics=True)
which is from a book (Python Machine Learning by @rasbt https://github.com/rasbt/python-machine-learning-book. It's awesome, but I guess the code gets old quick!)

The new way of doing it means to add metrics=['accuracy'] to the model.compilesection, rather than themodel.fit`:

model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])
I've just run into this issue, and changing the code in these places worked for me :D

[rasbt]

@JAStark The keras section? Oh yeah, that was 2015 :P. The next edition will have up-to-date tensorflow >= 1.0 code ;)

[JAStark]

@rasbt yes! Chapter 13 :)

[maxpumperla]

ok, it's updated now in the readme. Note that you're free to leave out additional arguments in examples as you want, hyperas is just a wrapper, not a prescription. :)

[beyhangl]

@JAStark thanks

[gontthias]

@JAStark good answer,thanks