while executing GMF.py

[mushahrukhkhan]

Hello,

I am getting this error while error GMF.Py kindly guide me....

ValueError: Only call sigmoid_cross_entropy_with_logits with named arguments (labels=..., logits=..., ...)

[NickShatalov]

Hi!

Got the same error. If someone could provide some help, I will really appreciate

[hexiangnan]

This seems to be an API discrepancy. Please check how is the use of the API in your Tensorflow/Theano version.

On Thu, Aug 16, 2018 at 12:03 AM, NickShatalov [email protected] wrote:

Hi!

Got the same error. If someone could provide some help, I will really appreciate

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/hexiangnan/neural_collaborative_filtering/issues/15#issuecomment-413244949, or mute the thread https://github.com/notifications/unsubscribe-auth/ABGxjtVgz2BRGPudkwMTyxdUUEjGvR_Tks5uREY9gaJpZM4S6vUn .

[NickShatalov]

Yes, this error was due to conflict of versions.

I had

• Keras 1.0.7
• Theano 0.8.2
• Tensorflow 1.9.0

In /home/user/anaconda3/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py I changed:

• tf.nn.sigmoid_cross_entropy_with_logits(output, target) to tf.nn.sigmoid_cross_entropy_with_logits(logits=output, labels=target),
• added from tensorflow.python import control_flow_ops as flow_ops with using flow_ops.cond(...) instead of tensorflow.python.control_flow_ops.cond(...)

Also I changed some old constructions in GMF.py:

• changed xramge to range
• changed train.has_key((u, j)) to (u, j) in train

After that model finally started to train

[rishj97]

This pull-request fixes the problems.

[Fanshaoliu]

Yes, this error was due to conflict of versions.

I had

• Keras 1.0.7
• Theano 0.8.2
• Tensorflow 1.9.0

In /home/user/anaconda3/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py I changed:

• tf.nn.sigmoid_cross_entropy_with_logits(output, target) to tf.nn.sigmoid_cross_entropy_with_logits(logits=output, labels=target),
• added from tensorflow.python import control_flow_ops as flow_ops with using flow_ops.cond(...) instead of tensorflow.python.control_flow_ops.cond(...)

Also I changed some old constructions in GMF.py:

• changed xramge to range
• changed train.has_key((u, j)) to (u, j) in train

After that model finally started to train

Thank U so much! I had fixed it