library_name: kernels
license: apache-2.0
tags:
- kernel
- webgpu
- wgsl
ai.onnx.Sum
ai.onnx · standard ONNX operator · ONNX opset ≥ 13
Description
Computes the elementwise sum of one or more input tensors with multidirectional (NumPy-style) broadcasting. All inputs and the output must share the same data type.
See the ONNX Sum spec for the reference semantics.
Inputs
| Name | Upstream name | Logical dtype | Rank | Shape | Description | Presence |
|---|---|---|---|---|---|---|
a |
A |
T |
— | — | First input tensor. | required |
b |
B |
T |
— | — | Second input tensor (optional). | optional |
c |
C |
T |
— | — | Third input tensor (optional). | optional |
d |
D |
T |
— | — | Fourth input tensor (optional). | optional |
Outputs
| Name | Upstream name | Logical dtype | Rank | Shape | Description | Presence |
|---|---|---|---|---|---|---|
y |
sum |
T |
derived | derived | Elementwise sum of all provided input tensors. | required |
Type constraints
| Variable | Allowed dtypes |
|---|---|
T |
float32, float16 |
Files
metadata.json— kernel metadata (id, digests, per-variant templates, provenance)manifest.json— the op contract (source of truth)test.json— correctness casesbench.json— benchmark + tuning casesdatamove-elementwise-copy.wgsl.jinjasummean-broadcast.wgsl.jinjasummean-vec4.wgsl.jinja
Use with @huggingface/kernels
npm install --save-exact @huggingface/kernels@0.0.1-preview.2
Required output shapes and logical data types are inferred from the supplied inputs and attributes; result tensors are allocated automatically.
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.Sum", { version: 1 });
const { y } = await kernel({ a: { data: aData, shape: [3] } });