# JAX v0.4.13 — JAX release v0.4.13

- Product: JAX (https://whatsnew.fyi/product/jax)
- Vendor: Google
- Date: 2023-06-23
- Version: v0.4.13
- Original notes: https://github.com/jax-ml/jax/releases/tag/jax-v0.4.13
- Permalink: https://whatsnew.fyi/product/jax/releases/v0.4.13

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

Reuse: the summaries, labels and curation here are © What's New. Quote freely with attribution and a link back; wholesale republication of the corpus is not permitted — terms: https://whatsnew.fyi/terms. The vendors' own release notes remain their publishers'.

---

- **changed** — jax.jit now allows None to be passed to in_shardings and out_shardings, with in_shardings marked as replicated and out_shardings determined by the XLA GSPMD partitioner
- **changed** — jax.experimental.pjit.pjit now allows None to be passed to in_shardings and out_shardings with mesh-dependent semantics
- **changed** — Executable.cost_analysis() now works on Cloud TPU
- **added** — Added a warning if a non-allowlisted jaxlib plugin is in use
- **added** — Added jax.tree_util.tree_leaves_with_path
- **fixed** — Fixed incorrect wheel name in CUDA 12 releases; the correct wheel is named cudnn89 instead of cudnn88
- **deprecated** — The native_serialization_strict_checks parameter to jax.experimental.jax2tf.convert is deprecated in favor of native_serializaation_disabled_checks

NOTE: This is the last JAX release that will include Python 3.8 support

* Changes
  * `jax.jit` now allows `None` to be passed to `in_shardings` and
    `out_shardings`. The semantics are as follows:
      * For in_shardings, JAX will mark is as replicated but this behavior
        can change in the future.
      * For out_shardings, we will rely on the XLA GSPMD partitioner to
        determine the output shardings.
  * `jax.experimental.pjit.pjit` also allows `None` to be passed to
    `in_shardings` and `out_shardings`. The semantics are as follows:
    * If the mesh context manager is *not* provided, JAX has the freedom to
      choose whatever sharding it wants.
      * For in_shardings, JAX will mark is as replicated but this behavior
        can change in the future.
      * For out_shardings, we will rely on the XLA GSPMD partitioner to
        determine the output shardings.
    * If the mesh context manager is provided, None will imply that the value
      will be replicated on all devices of the mesh.
  * Executable.cost_analysis() works on Cloud TPU
  * Added a warning if a non-allowlisted `jaxlib` plugin is in use.
  * Added `jax.tree_util.tree_leaves_with_path`.

* Bug fixes
  * Fixed incorrect wheel name in CUDA 12 releases (#16362); the correct wheel
    is named `cudnn89` instead of `cudnn88`.

* Deprecations
  * The `native_serialization_strict_checks` parameter to
    {func}`jax.experimental.jax2tf.convert` is deprecated in favor of the
    new `native_serializaation_disabled_checks` ({jax-issue}`#16347`).