Xenova HF Staff

sync 6fdf6301e2bb

b301fb9 verified 7 days ago

4.9 kB

library_name: kernels
license: apache-2.0
tags:
  - kernel
  - webgpu
  - wgsl

ai.onnx.ReduceProd

ai.onnx · standard ONNX operator · ONNX opset ≥ 18

Description

Computes the product of elements along specified axes of the input tensor. The output rank matches the input when keepdims is 1; reduced dimensions are pruned when keepdims is 0. Reduction over an empty set of values yields 1.

See the ONNX ReduceProd spec for the reference semantics.

Inputs

Name	Upstream name	Logical dtype	Rank	Shape	Description	Presence
`x`	`data`	`T`	—	—	The input tensor to reduce.	required

Outputs

Name	Upstream name	Logical dtype	Rank	Shape	Description	Presence
`y`	`reduced`	`T`	derived	—	The product-reduced output tensor.	required

Attributes

Default values (overridable per request):

Attribute	Default	Description
`axes`	`[]`	Values of the optional ONNX `axes` tensor input, supplied through this request attribute; an empty list follows `noop_with_empty_axes`.
`keepdims`	`1`	If 1, retains reduced dimensions with size 1 in the output; if 0, the reduced dimensions are removed.
`noop_with_empty_axes`	`0`	If 0 (default), empty or absent axes trigger reduction over all axes; if 1, empty axes make the op a no-op identity pass-through.

Type constraints

Variable	Allowed dtypes
`T`	`float32`, `float16`, `int32`

Implementation variants

One implementation is selected per call from the device capabilities, the request shapes and the dtypes; these notes say what each one covers.

axis0_splitk_i32 — Partitions a long rank-two axis-zero integer reduction across workgroups and combines exact int32 partials. It applies when the reduced row dimension is too large for one pass to expose enough parallelism.
subgroup_last_axis_vec4 — Reduces each contiguous last-axis row with subgroup collectives and vec4-packed reads.
subgroup_last_axis — Reduces each contiguous last-axis row with subgroup collectives and scalar reads for an unaligned row width.

Device requirements

Some implementation variants require subgroups. These are route-specific capabilities, not package-wide requirements; availability also depends on the request shape and dtype.

Files

metadata.json — kernel metadata (id, digests, per-variant templates, provenance)
manifest.json — the op contract (source of truth)
test.json — correctness cases
bench.json — benchmark cases
reduce-axis-split-reduce.wgsl.jinja
reduce-axis0-splitk-combine.wgsl.jinja
reduce-axis0-splitk-reduce.wgsl.jinja
reduce-axis0-tilecols.wgsl.jinja
reduce-flat-partial.wgsl.jinja
reduce-i32-axes02.wgsl.jinja
reduce-noop-empty-axes.wgsl.jinja
reduce-row-subgroup-rows.wgsl.jinja
reduce-row-subgroup.wgsl.jinja
reduce-row-tree.wgsl.jinja
reduce-serial-axis.wgsl.jinja

Use with `@huggingface/kernels`

npm install --save-exact @huggingface/kernels@0.0.1-preview.3

Outputs with inferable metadata are allocated automatically. Explicit outputs entries request optional results or provide metadata that cannot be inferred from the supplied inputs and attributes.

This example supplies explicit metadata for:

The version: 1 option selects the published kernel contract; it is independent of any operator opset, contrib since_version, or model version. It follows the v1 branch as fixes land. To pin exact artifact bytes, pass a 40-character commit revision instead of version.

Replace each *Data placeholder with a typed array containing the corresponding input data.

import { getKernel } from "@huggingface/kernels";

const kernel = await getKernel("webgpu-kernels/ai.onnx.ReduceProd", { version: 1 });
// Explicit destinations request optional results or supply metadata that cannot be inferred.
const { y } = await kernel({ x: { data: xData, shape: [3, 2, 2] } }, {
  outputs: { y: { shape: [1, 1, 1], dtype: "float32" } },
});