Source code
Revision control
Copy as Markdown
Other Tools
Test Info:
- This WPT test may be referenced by the following Test IDs:
- /webnn/conformance_tests/reduce_l2.https.any.html?cpu - WPT Dashboard Interop Dashboard
- /webnn/conformance_tests/reduce_l2.https.any.html?gpu - WPT Dashboard Interop Dashboard
- /webnn/conformance_tests/reduce_l2.https.any.html?npu - WPT Dashboard Interop Dashboard
- /webnn/conformance_tests/reduce_l2.https.any.worker.html?cpu - WPT Dashboard Interop Dashboard
- /webnn/conformance_tests/reduce_l2.https.any.worker.html?gpu - WPT Dashboard Interop Dashboard
- /webnn/conformance_tests/reduce_l2.https.any.worker.html?npu - WPT Dashboard Interop Dashboard
// META: title=test WebNN API reduction operations
// META: global=window,dedicatedworker
// META: variant=?cpu
// META: variant=?gpu
// META: variant=?npu
// META: script=../resources/utils.js
// META: timeout=long
'use strict';
// Reduce the input tensor along all dimensions, or along the axes specified in
// the axes array parameter.
//
// dictionary MLReduceOptions {
// sequence<[EnforceRange] unsigned long> axes;
// boolean keepDimensions = false;
// };
//
// MLOperand reduceL2(MLOperand input, optional MLReduceOptions options = {});
const getReductionOperatorsPrecisionTolerance = (graphResources) => {
return {
metricType: 'ULP',
value: getReducedElementCount(graphResources) * 2 + 1,
};
};
const reduceL2Tests = [
// reduceL2 tests
{
'name': 'reduceL2 float32 0D constant tensor default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [4.860228061676025],
'descriptor': {shape: [], dataType: 'float32'},
'constant': true
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 4.860228061676025,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 0D constant tensor empty axes',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [4.860228061676025],
'descriptor': {shape: [], dataType: 'float32'},
'constant': true
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}, {'options': {'axes': []}}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 4.860228061676025,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 1D constant tensor all positive default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [24], dataType: 'float32'},
'constant': true
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 1D tensor all positive default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [24], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 1D tensor all negative default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
-66.80043029785156, -53.00004959106445, -59.58587646484375,
-46.14392852783203, -49.60614013671875, -12.832738876342773,
-88.05061340332031, -75.56246185302734, -50.76777648925781,
-36.96630096435547, -26.344043731689453, -58.90546417236328,
-94.28752899169922, -22.7802791595459, -84.3487777709961,
-60.47734451293945, -41.455806732177734, -92.84781646728516,
-85.05448913574219, -30.235260009765625, -47.33808135986328,
-25.268428802490234, -78.11959075927734, -28.330944061279297
],
'descriptor': {shape: [24], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 292.57574462890625,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 1D tensor all positive integers default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4, 29, 8, 56, 42, 78, 89, 64, 56, 81, 85, 18,
6, 39, 35, 63, 87, 50, 81, 89, 5, 8, 37, 37
],
'descriptor': {shape: [24], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 274.4029846191406,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 1D tensor all negative integers default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
-70, -78, -65, -77, -25, -47, -63, -67, -66, -15, -28, -75,
-88, -54, -13, -27, -5, -18, -68, -71, -50, -56, -99, -99
],
'descriptor': {shape: [24], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 300.3830871582031,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 2D tensor default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [4, 6], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 3D tensor default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 3, 4], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 4D tensor default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 5D tensor default options',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 1, 4, 1, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 3D tensor options.axes',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 3, 4], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [{'input': 'reduceL2Input'}, {'options': {'axes': [2]}}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [
122.352783203125, 124.8213119506836, 128.20062255859375,
128.14801025390625, 87.18083953857422, 55.043975830078125
],
'descriptor': {shape: [2, 3], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 4D tensor options.axes',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments':
[{'input': 'reduceL2Input'}, {'options': {'axes': [0, 2]}}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [
114.44775390625, 110.26422882080078, 133.47344970703125,
64.96752166748047, 128.0914764404297, 101.677734375
],
'descriptor': {shape: [2, 3], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 3D tensor options.keepDimensions=false',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 3, 4], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [
{'input': 'reduceL2Input'}, {'options': {'keepDimensions': false}}
],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 3D tensor options.keepDimensions=true',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 3, 4], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments':
[{'input': 'reduceL2Input'}, {'options': {'keepDimensions': true}}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [272.0996398925781],
'descriptor': {shape: [1, 1, 1], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 4D tensor options.keepDimensions=false',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [
{'input': 'reduceL2Input'}, {'options': {'keepDimensions': false}}
],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': 272.0996398925781,
'descriptor': {shape: [], dataType: 'float32'}
}
}
}
},
{
'name': 'reduceL2 float32 4D tensor options.keepDimensions=true',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments':
[{'input': 'reduceL2Input'}, {'options': {'keepDimensions': true}}],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [272.0996398925781],
'descriptor': {shape: [1, 1, 1, 1], dataType: 'float32'}
}
}
}
},
{
'name':
'reduceL2 float32 4D tensor options.axes with options.keepDimensions=false',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [
{'input': 'reduceL2Input'},
{'options': {'axes': [1, 3], 'keepDimensions': false}}
],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [
138.580078125, 166.67791748046875, 149.91552734375, 67.6578598022461
],
'descriptor': {shape: [2, 2], dataType: 'float32'}
}
}
}
},
{
'name':
'reduceL2 float32 4D tensor options.axes with options.keepDimensions=true',
'graph': {
'inputs': {
'reduceL2Input': {
'data': [
4.860228061676025, 88.23184204101562, 54.489688873291016,
64.75027465820312, 6.855991363525391, 91.39871215820312,
41.88857650756836, 73.65444946289062, 35.31573486328125,
48.345428466796875, 82.39190673828125, 77.86200714111328,
93.31141662597656, 62.48688507080078, 60.29290008544922,
13.230599403381348, 20.535987854003906, 53.45161819458008,
11.320085525512695, 64.75763702392578, 43.6589469909668,
0.8374307155609131, 0.6848266124725342, 33.504703521728516
],
'descriptor': {shape: [2, 2, 2, 3], dataType: 'float32'}
}
},
'operators': [{
'name': 'reduceL2',
'arguments': [
{'input': 'reduceL2Input'},
{'options': {'axes': [1, 3], 'keepDimensions': true}}
],
'outputs': 'reduceL2Output'
}],
'expectedOutputs': {
'reduceL2Output': {
'data': [
138.580078125, 166.67791748046875, 149.91552734375, 67.6578598022461
],
'descriptor': {shape: [2, 1, 2, 1], dataType: 'float32'}
}
}
}
}
];
if (navigator.ml) {
reduceL2Tests.forEach((test) => {
webnn_conformance_test(
buildAndExecuteGraph, getReductionOperatorsPrecisionTolerance, test);
});
} else {
test(() => assert_implements(navigator.ml, 'missing navigator.ml'));
}