tensorflow / tensorflow/probability

JAX backend doesn't use `jax.tree_util`

Open
#1,785 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Jupyter Notebook
Stars
4.4k
Forks
1.1k
PR merge metrics
No merged PRs in 30d

Description

In this part, expand_composites is set to False and this prevents jax.tree_util for kicking in.
https://github.com/tensorflow/probability/blob/64a70d0850f2ab6b69693b7fb728b7c6eb759a7e/tensorflow_probability/python/internal/backend/numpy/nest.py#L314-L320

The resulting problem is that we cannot use pytree incompatible with dm_tree

Herę is notebook reproducing the bug
https://colab.research.google.com/gist/krzysztofrusek/a9fa71ca2bf3952a9f18358309225107/eqx_tfp.ipynb

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with tensorflow_probability/python/internal/backend/numpy/nest.py at lines 314-320 and run the linked Colab notebook to reproduce the incompatibility with dm_tree. Trace how expand_composites affects tree handling, then verify that the reported pytree case works without breaking the existing nest behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
backend, machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.