Hello hello,
Could I get some details about Keras Policy?
I mean, its inputs are just the stories?
And what can Keras Policy figure out from the stories? Checks all features of the current conversation, and ‘compares’ it with all states of all stories?