Now we will see how to perform a copy task using NTM. The goal of the copy task is to see how NTM stores and recalls a sequence of arbitrary length. We will feed the network a random sequence, along with a marker indicating the end of a sequence. It has to learn to output the given input sequence. So, the network will store the input sequence in the memory and then it will read back from the memory. Now, we will see step by step how to perform a copy task, and then we will see the final code as a whole at the end.
You can also check the code available as a Jupyter Notebook with an explanation here: https://github.com/sudharsan13296/Hands-On-Meta-Learning-With-Python/blob/master/05.%20Memory%20Augmented%20Networks/5.4%20Copy%20Task%20Using%20NTM.ipynb.
First, we will see how to implement the NTM cell. Instead of looking at the whole code, we will look at it...
