chore(deps): bump the rust-dependencies group across 1 directory with 4 updates

[dependabot]

Bumps the rust-dependencies group with 4 updates in the /src-tauri directory: ort, candle-core, candle-nn and safetensors.

Updates ort from 2.0.0-rc.12 to 2.0.0-rc.13

Release notes

Sourced from ort's releases.

v2.0.0-rc.13

2.0.0-rc.13

💖 If you find ort useful, please consider sponsoring us on Open Collective 💖

🤔 Need help upgrading? Ask questions in GitHub Discussions or in the pyke.io Discord server!

---

🎯 Execution provider clarity

EP structs like ep::CUDA are now compile-time gated behind their respective Cargo feature flags. The EP feature flags only modifying runtime behavior was a common pain point, and this change brings ort in line with, like, every other Rust crate in existence.

⚠️ ort will now throw an error at link time if download-binaries is enabled and no binaries contain all of the requested EPs. Previously, if you enabled a nonsensical combination like cuda + coreml, ort would silently fall back to a CPU-only build. This caused great confusion as the only indicator of what went wrong was in log messages that don't show by default. A new lax-feature-matching Cargo feature will fallback to the closest fit instead of erroring.

🍎 ONNX Runtime 1.28

rc.13 skips ahead 4 ONNX Runtime versions to v1.28, bringing new operator support, bug & security fixes, and performance improvements.

✖️ No CUDA 12

rc.13 only ships CUDA 13 binaries, as ONNX Runtime has deprecated CUDA 12.

🧩 Custom operator ergonomics

Custom operators have been reworked to have a more Rusty interface:

 struct MyOperator;
impl Operator for MyOperator {


type Kernel<'attr> = ort::operator::BoxedKernel<'attr>;
fn name(&self) -> &str {
"MyOperator"
}



fn inputs(&self) -> Vec<OperatorInput> {

  vec![OperatorInput::required(TensorElementType::Float32), OperatorInput::required(TensorElementType::Float32)]




fn inputs(&self) -> impl IntoIterator<Item = OperatorInput> {

  [OperatorInput::required(TensorElementType::Float32), OperatorInput::required(TensorElementType::Float32)]

}


fn outputs(&self) -> Vec<OperatorOutput> {

  vec![OperatorOutput::required(TensorElementType::Float32)]




fn outputs(&self) -> impl IntoIterator<Item = OperatorOutput> {

  [OperatorOutput::required(TensorElementType::Float32)]

}


fn create_kernel(&self, _: &KernelAttributes) -> ort::Result<Box<dyn Kernel>> {

  Ok(Box::new(|ctx: &amp;KernelContext| {



</tr></table>

... (truncated)

Commits

• 002f41a 2.0.0-rc.13
• d5dc28f chore(tract): update tract to v0.23
• 62ad710 chore(candle): update to candle v0.11.0
• 66ffef6 docs: update
• 58ebfd8 chore(tools): update api-coverage
• 71c1c84 refactor!: separate Device and HardwareDevice
• 85f4893 chore(sys): update ONNX Runtime to v1.28.0
• d7d924e refactor!: move environment::current to an associated function
• db48c03 feat: Tensor::size
• 2c6652b feat: per-run profiling
• Additional commits viewable in compare view

Updates candle-core from 0.9.2 to 0.10.2

Commits

• See full diff in compare view

Updates candle-nn from 0.9.2 to 0.10.2

Commits

• See full diff in compare view

Updates safetensors from 0.6.2 to 0.7.0

Release notes

Sourced from safetensors's releases.

v0.7.0

What's new

complex64 support

This release adds support for storing complex64 tensors.

What's Changed

• Revert "Early bailing when keys mismatch (faster). (#602)" by @​Narsil in huggingface/safetensors#647
• Support using paddle in safe_open by @​zeroRains in huggingface/safetensors#630
• Fix clippy lint. by @​Narsil in huggingface/safetensors#648
• Support paddle save/load/save_file/load_file without coverting to numpy by @​changeyoung98 in huggingface/safetensors#646
• Fix no-std feature configuration to auto-fallback from std to alloc by @​antimora in huggingface/safetensors#651
• Fix big endian (S390X) test by @​danieldk in huggingface/safetensors#654
• Fixing get_slice. by @​Narsil in huggingface/safetensors#655
• Update hashbrown dependency to 0.16 by @​danieldk in huggingface/safetensors#658
• Fix documentation build by @​danieldk in huggingface/safetensors#667
• Add packaging as a dependency for the torch extra by @​danieldk in huggingface/safetensors#666
• Support the complex64 data type by @​danieldk in huggingface/safetensors#660
• feat: add header size check at serialization by @​McPatate in huggingface/safetensors#669
• Set version to 0.7.0-dev.0 by @​danieldk in huggingface/safetensors#670
• Fix release CI issues by @​danieldk in huggingface/safetensors#673

New Contributors

• @​zeroRains made their first contribution in huggingface/safetensors#630
• @​changeyoung98 made their first contribution in huggingface/safetensors#646
• @​antimora made their first contribution in huggingface/safetensors#651
• @​danieldk made their first contribution in huggingface/safetensors#654

Full Changelog: safetensors/safetensors@v0.6.2...v0.7.0

Changelog

Sourced from safetensors's changelog.

Releasing safetensors

This document covers the full release process for the safetensors project. If anything here is unclear or out of date, please open a PR.

What gets released

A single tag push triggers two releases:

• Rust core crate on crates.io, via .github/workflows/rust-release.yml
• Python wheels + sdist on PyPI, via .github/workflows/python-release.yml (with build-provenance attestations)

Both workflows trigger on tags matching v* (e.g. v0.8.0, v0.8.0-rc.0).

Conda packages are published separately via the conda-forge/safetensors-feedstock, which picks up new PyPI releases automatically. No action is required from us.

The Rust crate and the Python binding share a version number. They could be versioned independently, but in practice they've always moved together.

Pre-release checklist

1. CI is green on main, every platform, every Python version.
2. Review the diff since the last release:

   git log --oneline v0.7.0..main
   git diff v0.7.0..main --stat

3. Identify breaking changes. Anything that changes the public Rust or Python API should be called out in the release notes and reflected in the version bump per semver.
4. Run benchmarks if the release includes performance-sensitive work:

   cd safetensors && cargo bench

5. Test against transformers. It's the largest downstream consumer of safetensors.
   • At minimum, run the fast test suite:

     RUN_PIPELINE_TESTS=1 CUDA_VISIBLE_DEVICES=-1 pytest -sv tests/

   • For any significant release, run the full suite by rebasing these two PRs on the transformers repo:
     • transformers#16708 — builds docker images with safetensors@main
     • transformers#16712 — runs the full test suite If those PRs have drifted, ask the transformers team for the current entry point.
6. Check the transformers version pin. transformers pins safetensors>=X,<Y in its setup.py. If this release crosses the upper bound, coordinate with the transformers maintainers first.

Cutting the release

We typically cut at least one release candidate before GA.

1. Update the dev version on main

main always carries a -dev.0 version. Before branching, make sure it reflects the upcoming release:

</tr></table> 

... (truncated)

Commits

• fa27fa8 Set version to 0.7.0
• 3979c8f Fix release CI issues (#673)
• 2624d79 Set version to 0.7.0-dev.0 (#670)
• 280141d feat: add header size check at serialization (#669)
• 547d0a7 Support the complex64 data type (#660)
• c6a98fa Add packaging as a dependency for the torch extra (#666)
• c7f11e3 Fix documentation build (#667)
• 2abf5c6 Update hashbrown dependency to 0.16 (#658)
• fa22263 Fixing get_slice. (#655)
• 34a3a27 Try to fix big endian test (#654)
• Additional commits viewable in compare view