I fit Stan models from time-to-time, but the optimization and sampling is pretty slow (for me), which is especially rough since I am often doing it many hundreds of thousands of times from the same model.
I am pretty sure I am leaving a lot on the table performance-wise, and that someone with a better knowledge of Stan could write the same model to run much faster. Rather than just post my code here every time and ask for free help to optimize it, I’d be open to paying the most expert Stan users/developers for their time to read the code and find the performance optimizations (or other issues).
I don’t know if there is a good platform for arranging this, or if this is even wanted (maybe the project leaders would prefer that I just request free help here each time). Any thoughts?