docs
docs copied to clipboard
Add tips on optimising performance for Turing models
~~Right now our only mention of TuringBenchmarking is in the AD bit~~
Would you consider expanding the scope of this issue to a more general issue on analysing performance problems, and fixing them? e.g. how do I figure out what bit of my model is slow, using standard Julia profiling tools? What do you, as the user, have control over?
TuringBenchmarking is removed, so I figure we should just change the scope of this to 'optimising performance'