I am trying to improve my approach to Deep Learning and could use some practical advice.
I am particularly interested in creative uses and customisation. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What trade-offs have you found...
