eqx-zoo 0.1.0__tar.gz

eqx_zoo-0.1.0/LICENSE

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

eqx_zoo-0.1.0/PKG-INFO

@@ -0,0 +1,100 @@
+Metadata-Version: 2.4
+Name: eqx-zoo
+Version: 0.1.0
+Summary: Verified Equinox ports of pretrained models, numerically matched against Hugging Face
+Keywords: jax,equinox,llm,transformers,pretrained-models,deep-learning
+Author: Zane
+Author-email: Zane <zane.n.neave@gmail.com>
+License-Expression: Apache-2.0
+License-File: LICENSE
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Typing :: Typed
+Requires-Dist: equinox>=0.12
+Requires-Dist: huggingface-hub>=0.30
+Requires-Dist: jax>=0.5
+Requires-Dist: jaxtyping>=0.3
+Requires-Dist: safetensors>=0.5
+Requires-Python: >=3.12
+Project-URL: Homepage, https://github.com/xquantize/eqx-zoo
+Project-URL: Repository, https://github.com/xquantize/eqx-zoo
+Project-URL: Issues, https://github.com/xquantize/eqx-zoo/issues
+Description-Content-Type: text/markdown
+
+<p align="center">
+  <img src="assets/banner.png" alt="eqx-zoo: verified Equinox ports of pretrained models" width="100%">
+</p>
+
+<p align="center">
+  <a href="https://github.com/xquantize/eqx-zoo/actions/workflows/ci.yml"><img src="https://github.com/xquantize/eqx-zoo/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
+  <a href="https://pypi.org/project/eqx-zoo/"><img src="https://img.shields.io/pypi/v/eqx-zoo" alt="PyPI"></a>
+  <a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache--2.0-blue" alt="License: Apache-2.0"></a>
+  <img src="https://img.shields.io/badge/python-3.12%2B-blue" alt="Python 3.12+">
+  <a href="https://github.com/patrick-kidger/equinox"><img src="https://img.shields.io/badge/built%20with-Equinox-8b5cf6" alt="Built with Equinox"></a>
+</p>
+
+eqx-zoo provides pretrained models as plain [Equinox](https://github.com/patrick-kidger/equinox) modules, loaded straight from Hugging Face checkpoints and numerically verified against the reference implementation in 🤗 Transformers.
+
+Every model is an ordinary pytree, so `jax.jit`, `jax.grad`, `jax.vmap` and the rest of the JAX ecosystem work on it directly.
+
+## Installation
+
+```bash
+pip install eqx-zoo
+```
+
+Requires Python 3.12+. The example below also uses `pip install tokenizers`.
+
+## Quick example
+
+```python
+import jax.numpy as jnp
+from tokenizers import Tokenizer
+
+from eqx_zoo import CausalLM, generate
+
+tokenizer = Tokenizer.from_pretrained("Qwen/Qwen3-0.6B")
+model = CausalLM.from_pretrained("Qwen/Qwen3-0.6B")
+
+prompt = jnp.array(tokenizer.encode("The capital of France is").ids)
+tokens = generate(model, prompt, max_new_tokens=30)
+print(tokenizer.decode(tokens.tolist()))
+# Paris. The capital of Italy is Rome. The capital of Spain is Madrid. ...
+```
+
+## Models
+
+Any checkpoint with a supported architecture loads with `CausalLM.from_pretrained`. These checkpoints are verified by the test suite:
+
+| Architecture | Verified checkpoints |
+|---|---|
+| `LlamaForCausalLM` | [SmolLM2-135M](https://huggingface.co/HuggingFaceTB/SmolLM2-135M), [Llama 3.2 1B](https://huggingface.co/meta-llama/Llama-3.2-1B) |
+| `Qwen3ForCausalLM` | [Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B) |
+| `Qwen2ForCausalLM` | [Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) |
+
+Each verified checkpoint is tested against Hugging Face activations captured in float32: every layer's output must match, and greedy generation must reproduce the reference output token for token. See [`tests/`](tests).
+
+## Development
+
+```bash
+git clone https://github.com/xquantize/eqx-zoo && cd eqx-zoo
+uv sync
+uv run python scripts/make_reference.py   # capture reference activations
+uv run pytest
+```
+
+Contributions are welcome. [`AGENTS.md`](AGENTS.md) describes the conventions every model follows.
+
+## See also
+
+- **Neural networks:** [Equinox](https://github.com/patrick-kidger/equinox).
+- **Typing:** [jaxtyping](https://github.com/patrick-kidger/jaxtyping).
+- **Other JAX model collections:** [Levanter](https://github.com/stanford-crfm/levanter) (LLM training), [Bonsai](https://github.com/jax-ml/bonsai) (Flax NNX), [Equimo](https://github.com/RhizomeResearch/Equimo) (vision).
+
+## License
+
+Apache 2.0. Pretrained weights are distributed by their original authors under their own licenses.

eqx_zoo-0.1.0/README.md

@@ -0,0 +1,73 @@
+<p align="center">
+  <img src="assets/banner.png" alt="eqx-zoo: verified Equinox ports of pretrained models" width="100%">
+</p>
+
+<p align="center">
+  <a href="https://github.com/xquantize/eqx-zoo/actions/workflows/ci.yml"><img src="https://github.com/xquantize/eqx-zoo/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
+  <a href="https://pypi.org/project/eqx-zoo/"><img src="https://img.shields.io/pypi/v/eqx-zoo" alt="PyPI"></a>
+  <a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache--2.0-blue" alt="License: Apache-2.0"></a>
+  <img src="https://img.shields.io/badge/python-3.12%2B-blue" alt="Python 3.12+">
+  <a href="https://github.com/patrick-kidger/equinox"><img src="https://img.shields.io/badge/built%20with-Equinox-8b5cf6" alt="Built with Equinox"></a>
+</p>
+
+eqx-zoo provides pretrained models as plain [Equinox](https://github.com/patrick-kidger/equinox) modules, loaded straight from Hugging Face checkpoints and numerically verified against the reference implementation in 🤗 Transformers.
+
+Every model is an ordinary pytree, so `jax.jit`, `jax.grad`, `jax.vmap` and the rest of the JAX ecosystem work on it directly.
+
+## Installation
+
+```bash
+pip install eqx-zoo
+```
+
+Requires Python 3.12+. The example below also uses `pip install tokenizers`.
+
+## Quick example
+
+```python
+import jax.numpy as jnp
+from tokenizers import Tokenizer
+
+from eqx_zoo import CausalLM, generate
+
+tokenizer = Tokenizer.from_pretrained("Qwen/Qwen3-0.6B")
+model = CausalLM.from_pretrained("Qwen/Qwen3-0.6B")
+
+prompt = jnp.array(tokenizer.encode("The capital of France is").ids)
+tokens = generate(model, prompt, max_new_tokens=30)
+print(tokenizer.decode(tokens.tolist()))
+# Paris. The capital of Italy is Rome. The capital of Spain is Madrid. ...
+```
+
+## Models
+
+Any checkpoint with a supported architecture loads with `CausalLM.from_pretrained`. These checkpoints are verified by the test suite:
+
+| Architecture | Verified checkpoints |
+|---|---|
+| `LlamaForCausalLM` | [SmolLM2-135M](https://huggingface.co/HuggingFaceTB/SmolLM2-135M), [Llama 3.2 1B](https://huggingface.co/meta-llama/Llama-3.2-1B) |
+| `Qwen3ForCausalLM` | [Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B) |
+| `Qwen2ForCausalLM` | [Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) |
+
+Each verified checkpoint is tested against Hugging Face activations captured in float32: every layer's output must match, and greedy generation must reproduce the reference output token for token. See [`tests/`](tests).
+
+## Development
+
+```bash
+git clone https://github.com/xquantize/eqx-zoo && cd eqx-zoo
+uv sync
+uv run python scripts/make_reference.py   # capture reference activations
+uv run pytest
+```
+
+Contributions are welcome. [`AGENTS.md`](AGENTS.md) describes the conventions every model follows.
+
+## See also
+
+- **Neural networks:** [Equinox](https://github.com/patrick-kidger/equinox).
+- **Typing:** [jaxtyping](https://github.com/patrick-kidger/jaxtyping).
+- **Other JAX model collections:** [Levanter](https://github.com/stanford-crfm/levanter) (LLM training), [Bonsai](https://github.com/jax-ml/bonsai) (Flax NNX), [Equimo](https://github.com/RhizomeResearch/Equimo) (vision).
+
+## License
+
+Apache 2.0. Pretrained weights are distributed by their original authors under their own licenses.

eqx_zoo-0.1.0/pyproject.toml

@@ -0,0 +1,92 @@
+[project]
+name = "eqx-zoo"
+version = "0.1.0"
+description = "Verified Equinox ports of pretrained models, numerically matched against Hugging Face"
+readme = "README.md"
+license = "Apache-2.0"
+license-files = ["LICENSE"]
+requires-python = ">=3.12"
+keywords = [
+    "jax",
+    "equinox",
+    "llm",
+    "transformers",
+    "pretrained-models",
+    "deep-learning",
+]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Science/Research",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Typing :: Typed",
+]
+dependencies = [
+    "equinox>=0.12",
+    "huggingface-hub>=0.30",
+    "jax>=0.5",
+    "jaxtyping>=0.3",
+    "safetensors>=0.5",
+]
+
+[[project.authors]]
+name = "Zane"
+email = "zane.n.neave@gmail.com"
+
+[project.urls]
+Homepage = "https://github.com/xquantize/eqx-zoo"
+Repository = "https://github.com/xquantize/eqx-zoo"
+Issues = "https://github.com/xquantize/eqx-zoo/issues"
+
+[build-system]
+requires = ["uv_build>=0.12.10,<0.13.0"]
+build-backend = "uv_build"
+
+[dependency-groups]
+dev = [
+    "pytest>=9.1.1",
+    "ruff>=0.16.9",
+    "torch>=2.14.0",
+    "transformers>=4.51.3",
+]
+
+[tool.uv]
+constraint-dependencies = [
+    "huggingface-hub<2",
+    "torch==2.14.0+cpu; sys_platform == 'linux' and platform_machine == 'x86_64'",
+]
+
+[[tool.uv.sources.torch]]
+index = "pytorch-cpu"
+marker = "sys_platform == 'linux'"
+
+[[tool.uv.index]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+explicit = true
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py312"
+
+[tool.ruff.lint]
+select = [
+    "E",
+    "F",
+    "I",
+    "UP",
+    "B",
+    "D",
+]
+ignore = ["F722"]
+
+[tool.ruff.lint.pydocstyle]
+convention = "google"
+
+[tool.ruff.lint.per-file-ignores]
+"tests/**" = ["D"]

eqx_zoo-0.1.0/pyproject.toml.orig

@@ -0,0 +1,79 @@
+[project]
+name = "eqx-zoo"
+version = "0.1.0"
+description = "Verified Equinox ports of pretrained models, numerically matched against Hugging Face"
+readme = "README.md"
+license = "Apache-2.0"
+license-files = ["LICENSE"]
+authors = [{ name = "Zane", email = "zane.n.neave@gmail.com" }]
+requires-python = ">=3.12"
+keywords = ["jax", "equinox", "llm", "transformers", "pretrained-models", "deep-learning"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Science/Research",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Typing :: Typed",
+]
+dependencies = [
+    "equinox>=0.12",
+    "huggingface-hub>=0.30",
+    "jax>=0.5",
+    "jaxtyping>=0.3",
+    "safetensors>=0.5",
+]
+
+[project.urls]
+Homepage = "https://github.com/xquantize/eqx-zoo"
+Repository = "https://github.com/xquantize/eqx-zoo"
+Issues = "https://github.com/xquantize/eqx-zoo/issues"
+
+[build-system]
+requires = ["uv_build>=0.12.10,<0.13.0"]
+build-backend = "uv_build"
+
+[dependency-groups]
+dev = [
+    "pytest>=9.1.1",
+    "ruff>=0.16.9",
+    "torch>=2.14.0",
+    "transformers>=4.51.3",
+]
+
+[tool.uv]
+constraint-dependencies = [
+    # Released transformers versions require huggingface-hub < 2. This only affects our
+    # lockfile; it is not part of the published package metadata.
+    "huggingface-hub<2",
+    # On the PyTorch CPU index, x86-64 Linux wheels are versioned `X.Y.Z+cpu`, and plain
+    # `X.Y.Z` has no wheel for that platform. Pin the variant so that
+    # `--resolution lowest-direct` can't select an uninstallable one.
+    "torch==2.14.0+cpu; sys_platform == 'linux' and platform_machine == 'x86_64'",
+]
+
+[tool.uv.sources]
+torch = [{ index = "pytorch-cpu", marker = "sys_platform == 'linux'" }]
+
+[[tool.uv.index]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+explicit = true
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py312"
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "UP", "B", "D"]
+ignore = ["F722"]  # jaxtyping shape strings like "seq dim" aren't Python expressions
+
+[tool.ruff.lint.pydocstyle]
+convention = "google"
+
+[tool.ruff.lint.per-file-ignores]
+"tests/**" = ["D"]

eqx_zoo-0.1.0/src/eqx_zoo/__init__.py

@@ -0,0 +1,11 @@
+"""Verified Equinox ports of pretrained models, numerically matched against Hugging Face."""
+
+from importlib.metadata import version
+
+from eqx_zoo.causal_lm import CausalLM
+from eqx_zoo.config import Config
+from eqx_zoo.generate import generate
+from eqx_zoo.layers import Cache
+
+__all__ = ["Cache", "CausalLM", "Config", "generate"]
+__version__ = version("eqx-zoo")

eqx_zoo-0.1.0/src/eqx_zoo/_loading.py

@@ -0,0 +1,67 @@
+"""Load safetensors checkpoints into Equinox modules by parameter name."""
+
+from collections.abc import Iterable
+from contextlib import ExitStack
+from pathlib import Path
+
+import jax
+import jax.numpy as jnp
+from safetensors import safe_open
+
+
+def _path_to_name(path: tuple) -> str:
+    parts = []
+    for key in path:
+        if isinstance(key, jax.tree_util.GetAttrKey):
+            parts.append(key.name)
+        elif isinstance(key, jax.tree_util.SequenceKey):
+            parts.append(str(key.idx))
+        else:
+            raise TypeError(f"unsupported pytree key {key!r}")
+    return ".".join(parts)
+
+
+def load_safetensors[T](
+    skeleton: T, files: Iterable[str | Path], *, ignore: Iterable[str] = ()
+) -> T:
+    """Fill a model skeleton with tensors from safetensors files.
+
+    Each leaf of `skeleton` is looked up by its attribute path, e.g. the leaf at
+    `model.layers[0].self_attn.q_proj.weight` is loaded from the checkpoint tensor
+    named `"model.layers.0.self_attn.q_proj.weight"`, and cast to the leaf's dtype.
+
+    Args:
+        skeleton: Module whose array leaves describe the expected shapes and dtypes,
+            typically created with `equinox.filter_eval_shape`.
+        files: Checkpoint files; parameters may be sharded across several.
+        ignore: Checkpoint parameters that are deliberately unused.
+
+    Returns:
+        `skeleton` with every array leaf replaced by the corresponding tensor.
+
+    Raises:
+        KeyError: If the checkpoint is missing a parameter.
+        ValueError: If a shape mismatches, or the checkpoint has unexpected parameters.
+    """
+    leaves, treedef = jax.tree_util.tree_flatten_with_path(skeleton)
+    names = [_path_to_name(path) for path, _ in leaves]
+
+    with ExitStack() as stack:
+        handles = [stack.enter_context(safe_open(f, framework="flax")) for f in files]
+        index = {name: handle for handle in handles for name in handle.keys()}
+
+        loaded = []
+        for name, (_, leaf) in zip(names, leaves, strict=True):
+            if name not in index:
+                raise KeyError(f"checkpoint is missing parameter {name!r}")
+            tensor = index[name].get_tensor(name)
+            if tuple(tensor.shape) != tuple(leaf.shape):
+                raise ValueError(f"{name}: checkpoint shape {tensor.shape} != model {leaf.shape}")
+            loaded.append(jnp.asarray(tensor, dtype=leaf.dtype))
+
+    unexpected = sorted(set(index) - set(names) - set(ignore))
+    if unexpected:
+        raise ValueError(
+            f"checkpoint has {len(unexpected)} unexpected parameters: {unexpected[:5]}"
+        )
+    return jax.tree_util.tree_unflatten(treedef, loaded)