Mostafa Samir

Impressive work! I'm certainly curious about how it was able to generalize with the same amount of memory locations! What do you think about taking it up a notch? Let's...

Great work Samu @Zeta36 ! **Regarding the adding task** I have a comment about how you apply the wights to the loss. You use the following: ```python loss = tf.reduce_mean(tf.square((loss_weights...

Yes, it should be an array of sequence lengths, that's indeed a current limitation in the implementation, and it's mentioned explicitly in the [usage guide](https://github.com/Mostafa-Samir/DNC-tensorflow/blob/master/docs/basic-usage.md#using-the-dnc-module) and implicitly in the [to-dos](https://github.com/Mostafa-Samir/DNC-tensorflow#to-do)....

Thanks for your contribution! Great job! But don't you think it would be better (form a UX perspective) to have a restart button when the game ends instead of automatically...

I tried it with netbeans and it was able to detect the library's APIs. I don't know about other IDEs but I have no reason to think it won't work...

yes, it can recognize classes made with Classing{js} and can enumerate their public properties but it can't enumerate the public properties in an instantiated object in that class I'm not...