Add Helion as a supported Python dependency

[sayakpaul]

What does this PR do?

Fixes #686

Adds Helion to the supported python-depends list so that Helion kernels can be built and published as noarch kernels:

[general]
backends = ["cuda", "rocm", "xpu"]
python-depends = ["helion"]

[torch-noarch]

Helion kernels are pure Python (Triton-based, JIT-compiled at runtime, with optional pretuned/AOT heuristics shipped as .py files next to the kernel source), so no new build machinery is needed — only a dependency-registry entry.

Verification

Tested end-to-end with a Helion vector-add kernel (based on Helion's pretuned vector_add) declaring backends = ["cuda", "rocm", "xpu"] and python-depends = ["helion"]:

• Bundle evaluates to all three noarch variants (torch-cuda, torch-rocm, torch-xpu) on x86_64-linux; without the overlay override, ROCm/XPU evaluation fails on the nixpkgs broken flag.
• The torch-cuda variant builds on aarch64-linux, passes get-kernel-check (which imports the kernel, and therefore Helion, inside the sandbox), and records "python-depends": ["helion"] in metadata.json.
• The built variant loads through kernels.get_local_kernel and produces correct results on an NVIDIA GB10.
• Helion's pretuned/AOT flow (helion.experimental.aot_kernel + _helion_aot_*_sm<NN>.py heuristic files, available since Helion 1.0.0) also works from a package dir: on sm121 it falls back to the sm100 heuristics and runs correctly.

Notes

• Helion requires torch >= 2.9 but does not declare it as a dependency (peer dep); the builder's Torch range satisfies this.
• nixpkgs is at Helion 1.0.0 while PyPI is at 1.2.0; if newer features are ever needed at build-check time, the overlay can bump the version.

Generated kernel: https://huggingface.co/kernels/sayakpaul/vector-add-helion

🤖 Generated with Claude Code. Edited by me (Sayak)

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[github-actions]

Coverage report — kernels/

Measured on: Python 3.10 / Torch 2.12.0.
Other CI configurations are not included in this number.
Hardware-gated code paths (ROCm/XPU/NPU/Darwin/Windows) are excluded or unreachable on the Linux+CUDA runner.

Total coverage: 84.6% — threshold: 80% — ✅

Per-file breakdown

Name	Stmts	Miss	Cover	Missing
src/kernels/__init__.py	10	0	100%
src/kernels/_system.py	6	1	83%	10
src/kernels/_versions.py	63	7	89%	46, 49, 52-53, 56-57, 100
src/kernels/backends.py	194	55	72%	40, 44, 48-51, 68, 90, 108, 117, 121, 125-127, 148, 170, 181, 188-191, 201, 205-225, 233, 256-276
src/kernels/compat.py	8	1	88%	5
src/kernels/deps.py	54	4	93%	58-59, 95, 98
src/kernels/layer/__init__.py	6	0	100%
src/kernels/layer/_interval_tree.py	103	4	96%	23, 52, 147, 150
src/kernels/layer/device.py	48	14	71%	42, 47-49, 91, 96-98, 101, 149, 152, 155-157
src/kernels/layer/func.py	81	7	91%	81, 111, 183, 301, 307, 320, 338
src/kernels/layer/globals.py	5	0	100%
src/kernels/layer/kernelize.py	74	8	89%	255, 281, 289-290, 296, 300, 316-318
src/kernels/layer/layer.py	210	16	92%	167, 210, 216, 229, 337, 417-418, 430, 439, 447, 458, 487, 491, 504, 557, 587
src/kernels/layer/mode.py	14	0	100%
src/kernels/layer/repos.py	130	34	74%	27, 33, 36-41, 61-62, 68, 71-74, 88, 92, 101-102, 108, 111-114, 121-122, 128, 131-134, 141-142, 148, 151-154, 235
src/kernels/lockfile.py	71	46	35%	37-104, 108-131
src/kernels/status.py	49	2	96%	23, 81
src/kernels/utils.py	301	55	82%	65, 77-81, 87-88, 218, 222, 225, 287, 295, 334-335, 373, 404, 409, 444, 673, 676, 678, 684, 697-698, 719-731, 735-742, 750, 754-764, 768-775, 813, 817, 836, 838
src/kernels/variants.py	262	19	93%	56, 87, 108, 138, 247-248, 289, 291, 371-378, 384-390, 421-427, 439-445, 534-536
src/kernels/verify.py	88	1	99%	32
TOTAL	1777	274	85%

Updated by the Test kernels workflow on commit 2839c9b08a5afc85629de024ce9e7f133b9c1181.

[sayakpaul]

Example kernel: https://huggingface.co/kernels/sayakpaul/vector-add-helion

Try it:

from kernels import get_kernel

kernel_module = get_kernel("sayakpaul/vector-add-helion", version=1)
vector_add = kernel_module.vector_add

vector_add(...)

[drbh]

looks good to me! (rebased and waiting to see if CI is green now)