PyMC4: custom likelihood

[Aravind27]

Is custom Aesara Ops still the best approach for this? I am reviving this discussion since it has been a while since the last comment. I am referring to the following comment by @michaelosthege

@michaelosthege

It sounds like you want to use a custom logp/custom likelihood which you can do using pm.Potential.
For that you can make the necessary computations with pure Aesara Ops if you want them to be differentiable, or you can implement a custom Aesara Op with or without a grad() implementation.

I would like to define a custom log likelihood and also provide its gradient to the NUTS sampler. Could someone kindly guide me on the best way to do this? Some examples of similar implementations would also be very helpful.

Thanks a lot for your time!