Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This builds on @aseyboldt's work in #944. It is part of the INLA hackathon (pymc-devs/pymc-experimental#340, specific related issues pymc-devs/pymc-experimental#342 and pymc-devs/pymc-experimental#343) and will later involve rewrites for the gradient using the adjoint method where we define the backward pass only at the solution point rather than backpropagating through each step of the iterator and storing all the intermediate values (implicit function theorem, see here, here, and here).
There's a crappy test that checks it matches the jax output from an example from the Deep Implicit Layers tutorial. Note, it's essential to run this with jax float64 precision
Related Issue
find_MAP
pymc#7308Checklist
Type of change