Run a forward pass through the given model

[stoplime]

The existing code did not have any way to generate an output from the network.

• Make a save of the model after training
• Figure out how the input of the model is setup
• Run a forward pass

[stoplime]

The input to the model is a feed_dict where the input is already translated into a dictionary index.

[stoplime]

It has been translated but there is an issue with the labels not being aligned properly.