MXFusion
MXFusion copied to clipboard
Add details about the machinery in Inference
The class structure is well documented but the details of how it all goes down is less so.
Such topics as:
- apply transformations to variable parameters at the beginning before execution and invert the transform when you ask for it.
- Works by essentially making a whole big gluon block that computes everything - the whole chain
- Where do parameters come from? how do model parameters vs posterior parameters vs gluon block parameters?