| --- |
| library_name: kernels |
| license: apache-2.0 |
| tags: |
| - kernel |
| - webgpu |
| - wgsl |
| --- |
| # ai.onnx.QLinearConv |
|
|
| `ai.onnx` · standard ONNX operator · ONNX opset ≥ 10 |
|
|
| ## Description |
|
|
| Performs convolution on a quantized input using quantized weights, producing a quantized output. Input and output quantization use per-tensor scales and zero points; weight quantization may be per-tensor or per-output-channel; an optional int32 bias must be pre-quantized with `scale = x_scale * w_scale` and zero point 0. |
|
|
| See the [ONNX `QLinearConv` spec](https://onnx.ai/onnx/operators/onnx__QLinearConv.html) for the reference semantics. |
|
|
| ## Inputs |
|
|
| | Name | Logical dtype | Rank | Shape | Description | Presence | |
| | --- | --- | --- | --- | --- | --- | |
| | `x` | `TX` | — | — | Quantized input data tensor shaped `(N x C x D1 x ... x Dn)`. | required | |
| | `x_scale` | `TF` | — | — | Per-tensor scale for input `x`. | required | |
| | `x_zero_point` | `TX` | — | — | Per-tensor zero point for input `x`. | required | |
| | `w` | `TW` | — | — | Quantized weight tensor shaped `(M x C/group x k1 x ... x kn)`. | required | |
| | `w_scale` | `TF` | — | — | Scale for weights `w`; scalar for per-tensor or 1-D of length M for per-output-channel quantization. | required | |
| | `w_zero_point` | `TW` | — | — | Zero point for weights `w`; scalar or 1-D of length M matching `w_scale`. | required | |
| | `y_scale` | `TF` | — | — | Per-tensor scale for output `y`. | required | |
| | `y_zero_point` | `TY` | — | — | Per-tensor zero point for output `y`. | required | |
| | `B` | `int32` | `1` | — | Optional 1-D bias of length `M`, pre-quantized with scale `x_scale * w_scale` and zero point 0. | optional | |
|
|
| ## Outputs |
|
|
| | Name | Logical dtype | Rank | Shape | Description | Presence | |
| | --- | --- | --- | --- | --- | --- | |
| | `y` | `TY` | same as `x` | derived | Quantized output tensor; shape is determined by kernel size, strides, and padding. | required | |
|
|
| ## Attributes |
|
|
| Attributes and default values (overridable per request): |
|
|
| | Attribute | Default | Description | |
| | --- | --- | --- | |
| | `auto_pad` | `"NOTSET"` | Automatic padding mode. `NOTSET` uses `pads`; `SAME_UPPER` and `SAME_LOWER` choose padding so each output spatial size is `ceil(input / stride)`; `VALID` uses no padding. | |
| | `dilations` | — | Optional dilation factors, one positive integer per spatial axis. Omission means all ones. | |
| | `group` | `1` | Number of groups that input and output channels are split into; defaults to 1. | |
| | `kernel_shape` | — | Optional kernel shape, one positive integer per spatial axis. When present, it must match the spatial dimensions of the weight tensor; omission infers the shape from the weights. | |
| | `pads` | — | Optional explicit padding in ONNX order `[begin_axis_0, ..., begin_axis_n, end_axis_0, ..., end_axis_n]`. Omission means all zeros; it cannot be combined with an automatic padding mode. | |
| | `strides` | — | Optional stride factors, one positive integer per spatial axis. Omission means all ones. | |
|
|
| ## Type constraints |
|
|
| | Variable | Allowed dtypes | |
| | --- | --- | |
| | `TX` | `uint8`, `int8` | |
| | `TW` | `uint8`, `int8` | |
| | `TY` | `uint8`, `int8` | |
| | `TF` | `float32` | |
|
|
| ## Files |
|
|
| - [`metadata.json`](build/webgpu/metadata.json) — kernel metadata (id, digests, per-variant templates, provenance) |
| - [`manifest.json`](build/webgpu/manifest.json) — the op contract (source of truth) |
| - [`test.json`](build/webgpu/test.json) — correctness cases |
| - [`bench.json`](build/webgpu/bench.json) — benchmark cases |
| - [`conv-int-accumulate-spatial.wgsl.jinja`](build/webgpu/conv-int-accumulate-spatial.wgsl.jinja) |
| - [`conv-int-im2col-spatial.wgsl.jinja`](build/webgpu/conv-int-im2col-spatial.wgsl.jinja) |
| - [`qlinear-conv-dp4a-direct-3x3.wgsl.jinja`](build/webgpu/qlinear-conv-dp4a-direct-3x3.wgsl.jinja) |
| - [`qlinear-conv-nchw-x4.wgsl.jinja`](build/webgpu/qlinear-conv-nchw-x4.wgsl.jinja) |
| - [`qlinear-conv-requantize.wgsl.jinja`](build/webgpu/qlinear-conv-requantize.wgsl.jinja) |
| - [`quant-dp4a-matmul.wgsl.jinja`](build/webgpu/quant-dp4a-matmul.wgsl.jinja) |
| - [`quant-pad-k-matrices.wgsl.jinja`](build/webgpu/quant-pad-k-matrices.wgsl.jinja) |
|
|
| ## Use with `@huggingface/kernels` |
|
|
| ```sh |
| npm install --save-exact @huggingface/kernels@0.0.1-preview.3 |
| ``` |
|
|
| 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. |
|
|
| ```js |
| import { getKernel } from "@huggingface/kernels"; |
| |
| const kernel = await getKernel("webgpu-kernels/ai.onnx.QLinearConv", { version: 1 }); |
| const { y } = await kernel({ |
| x: { data: xData, shape: [1, 1, 1] }, |
| x_scale: { data: x_scaleData, shape: [1] }, |
| x_zero_point: { data: x_zero_pointData, shape: [1] }, |
| w: { data: wData, shape: [1, 1, 1] }, |
| w_scale: { data: w_scaleData, shape: [1] }, |
| w_zero_point: { data: w_zero_pointData, shape: [1] }, |
| y_scale: { data: y_scaleData, shape: [1] }, |
| y_zero_point: { data: y_zero_pointData, shape: [1] }, |
| }); |
| ``` |
|
|
