ai.onnx.AveragePool

ai.onnx · standard ONNX operator · ONNX opset ≥ 19

Description

Applies average pooling over a sliding kernel window on input tensor X, computing the mean of values within each window position and writing results to Y. Output spatial dimensions are determined by kernel_shape, strides, dilations, pads, and ceil_mode; padded positions are excluded from the average by default unless count_include_pad is set.

See the ONNX AveragePool spec for the reference semantics.

Inputs

Name	Bind key	Logical dtype	Rank	Shape	Description	Presence
`X`	`x`	`T`	—	—	Input data tensor of shape `(N x C x D1 x D2 ... Dn)`, where `N` is the batch size and `C` is the number of channels.	required

Outputs

Name	Bind key	Logical dtype	Rank	Shape	Description	Presence
`Y`	`y`	`T`	same as `X`	derived; see description	Output tensor with pooled values; spatial dimensions vary based on kernel, stride, dilation, and pad settings.	required

Attributes

Attributes and default values (overridable per request):

Attribute	Default	Description
`auto_pad`	`"NOTSET"`	Deprecated auto-padding mode: `NOTSET` (use explicit pads), `SAME_UPPER`, `SAME_LOWER` (pad so output size is `ceil(input / stride)`), or `VALID` (no padding).
`count_include_pad`	`0`	When non-zero, pad pixels are counted in the divisor when computing the average; defaults to 0 (exclude pad).
`ceil_mode`	`0`	When non-zero, uses ceiling instead of floor when computing the output spatial shape; defaults to 0.
`kernel_shape`	—	Required kernel shape, with one positive value per spatial axis.
`strides`	—	Stride along each spatial axis. When omitted, every stride is 1.
`pads`	—	Padding at the beginning and end of each spatial axis, ordered as `[begin_0, ..., begin_n, end_0, ..., end_n]`. When omitted, every pad is 0.
`dilations`	—	Dilation along each spatial axis. When omitted, every dilation is 1.

Type constraints

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

Files

metadata.json — kernel metadata (id, digests, provenance)
manifest.json — the op contract (source of truth)
test.json — correctness cases
bench.json — benchmark + tuning cases
average-pool2d-nchw-horizontal-reuse.wgsl.jinja
average-pool2d-nchw-w3s1-reuse.wgsl.jinja
pool-global-reduction.wgsl.jinja
pool-window-nd.wgsl.jinja
pool-window-unroll.wgsl.jinja

Use with `@huggingface/kernels`

The loader derives every required output's shape and logical dtype from the manifest contract and this call. It then allocates the result tensors automatically.

The version: 1 option selects the published kernel contract; it is independent of any operator opset, contrib since_version, or model 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.AveragePool", { version: 1 });
const { y } = await kernel({ x: { data: xData, shape: [1, 3, 32] } }, {
  attrs: { kernel_shape: [2] },
});

Downloads last month: -

kernel

webgpu

wgsl

apache-2.0

WebGPU

Requires WebGPU support. See the compatibility table.