-
NumPyro is an excellent package that offers a user-friendly and minimal interface for defining models. However, the added abstraction can lead to increased memory usage and slower processing times. To…
-
I'm assuming these functions are expected to work with seeded models, but they yield ``UnexpectedTracerError``s.
```python
import numpyro
import jax
import jax.random as jr
import numpyro.distrib…
-
### Describe the issue:
When I use the blackjax backend, I get datetype errors.
the code runs fine with the pymc and nutpie samplers.
### Reproduceable code example:
```python
import numpy as np…
-
### Current behavior
Blackjax doesn't have code for constraints.
Stan (and NumPyro) handle constraints by transformation, and one could extract the Numpyro transformations and use those.
…
-
Our `RandomVariable`s which use `scan` cannot be used in a `numpyro.plate` context (see https://num.pyro.ai/en/stable/primitives.html#scan). However, we can still write reasonable looking models by us…
-
See https://github.com/arviz-devs/arviz/issues/2367 for example. We should look into what makes it be slow and try to fix it, it would also be nice to add a progressbar so it is less surprising when t…
-
-
Jax-0.4.31: Runtime: 27.06 seconds
https://colab.research.google.com/drive/1EsFY1St8Y2ZNBZ9UXTa9FDWrjPDdTU4U?usp=sharing
Jax-0.4.33: Runtime: 84.91 seconds
https://colab.research.google.com/drive…
-
Hi there,
Thank you very much for putting this package together, this is impressive! I was wondering if you would be interested in an implementation of nested sampling in pure jax. I know that the […
-
Hello again,
I've been trying to sample the posterior as done in your examples, and I was wondering if there was a way to have the chains run in parallel? I got thrown a warning saying:
"UserW…