About Anomaly Detection Template for TIBCO Spotfire v4.0.0.dxp

[nttcom online]

Sorry for asking so many questions.

When I click "Creat Model" on the "2-Model" page, I get the error "Could not execute function call'Autoencoder (Tensorflow / Keras)".

Looking at the contents of the error, I see the errors "name 'fitted_model' is not defined" and "Error in compiling Keras model: Could not interpret optimizer identifier: ".

I haven't changed anything and just run it, but I get above error.

If anyone knows the solution, please let me know.

[Sweta Kotha]

Hello, this has to do with compatibility issues across different Python package versions and passing in TensorFlow optimizers. If you could please retry with this updated DXP (tested on TensorFlow 2.9.1) and let us know if any issues arise.

[Sweta Kotha]

Hello, this has to do with compatibility issues across different Python package versions and passing in TensorFlow optimizers. If you could please retry with this updated DXP (tested on TensorFlow 2.9.1) and let us know if any issues arise.

[Marcus Horbach]

Using

optimizer=tf.keras.optimizers.Adam(learning_rate=learning_rate)'

is not compatible with

from tensorflow.python.keras import Sequential, Model

from tensorflow.python.keras.layers import Dense, Dropout

When I use this:

from tensorflow.keras import Sequential, Model

from tensorflow.keras.layers import Dense, Dropout

then the compile and fit work fine.

[Marcus Horbach]

Using

optimizer=tf.keras.optimizers.Adam(learning_rate=learning_rate)'

is not compatible with

from tensorflow.python.keras import Sequential, Model

from tensorflow.python.keras.layers import Dense, Dropout

When I use this:

from tensorflow.keras import Sequential, Model

from tensorflow.keras.layers import Dense, Dropout

then the compile and fit work fine.

[David Katz]

Thanks very much for your efforts. I'm working on this now and should have some answers soon.

[David Katz]

Thanks very much for your efforts. I'm working on this now and should have some answers soon.

[David Katz]

Thanks very much for your interest and your efforts. I'm working on this now and should have some answers soon.

 

It appears that in your video, the model training did not complete. The output path appears to contain some abnormal characters, which may have caused this. You could try blanking out that field and rerunning. Clear the error messages first so you can see only new ones.

 

Also, I'm working on some other fixes and should have a new version ready by tomorrow.

 

Regards,