// META: title=test WebNN API element-wise greater operation
// META: global=window,dedicatedworker
// META: variant=?cpu
// META: variant=?gpu
// META: variant=?npu
// META: script=../resources/utils.js
// META: timeout=long
'use strict';
// https://www.w3.org/TR/webnn/#api-mlgraphbuilder-logical
// Compare if the values of the first input tensor is greater, element-wise.
//
// MLOperand greater(MLOperand a, MLOperand b);
const getGreaterPrecisionTolerance = (graphResources) => {
const toleranceValueDict = {uint8: 0};
const expectedDataType =
getExpectedDataTypeOfSingleOutput(graphResources.expectedOutputs);
return {metricType: 'ULP', value: toleranceValueDict[expectedDataType]};
};
const greaterTests = [
{
'name': 'greater float32 0D scalar',
'graph': {
'inputs': {
'inputA': {
'data': [3.6851015090942383],
'descriptor': {'dimensions': [], 'dataType': 'float32'}
},
'inputB': {
'data': [1.723199725151062],
'descriptor': {'dimensions': [], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output':
{'data': [1], 'descriptor': {'dimensions': [], 'dataType': 'uint8'}}
}
}
},
{
'name': 'greater float32 1D constant tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [24], 'dataType': 'float32'},
'constant': true
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [24], 'dataType': 'float32'},
'constant': true
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [24], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 1D tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [24], 'dataType': 'float32'}
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [24], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [24], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 2D tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [4, 6], 'dataType': 'float32'}
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [4, 6], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [4, 6], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 3D tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 3, 4], 'dataType': 'float32'}
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [2, 3, 4], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [2, 3, 4], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 4D tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 5D tensors',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 1, 2, 3], 'dataType': 'float32'}
},
'inputB': {
'data': [
-6.155234336853027, -4.023341178894043, 5.9525980949401855,
2.306623697280884, -2.7692291736602783, -0.9711201190948486,
1.222206711769104, 4.590261459350586, 9.101232528686523,
-4.997007846832275, -4.80729341506958, 8.919360160827637,
0.9005027413368225, -2.8414556980133057, -2.8280413150787354,
8.47984504699707, -7.84067964553833, 9.213960647583008,
4.982365131378174, -2.507319211959839, -4.518013954162598,
8.351094245910645, -6.161073207855225, 0.7364829182624817
],
'descriptor': {'dimensions': [2, 2, 1, 2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0
],
'descriptor': {'dimensions': [2, 2, 1, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 broadcast 0D to 4D',
'graph': {
'inputs': {
'inputA': {
'data': [6.2216410636901855],
'descriptor': {'dimensions': [], 'dataType': 'float32'}
},
'inputB': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 broadcast 1D to 4D',
'graph': {
'inputs': {
'inputA': {
'data': [6.2216410636901855],
'descriptor': {'dimensions': [1], 'dataType': 'float32'}
},
'inputB': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 broadcast 2D to 4D',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
},
'inputB': {
'data': [
-2.684664487838745, 6.170023441314697, 9.487744331359863,
-2.5556411743164062, -2.0436434745788574, 8.533930778503418
],
'descriptor': {'dimensions': [2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 1, 0,
1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 broadcast 3D to 4D',
'graph': {
'inputs': {
'inputA': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
},
'inputB': {
'data': [
-7.099076271057129, -7.781408309936523, 8.782817840576172,
-8.948624610900879
],
'descriptor': {'dimensions': [2, 2, 1], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1,
1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
},
{
'name': 'greater float32 broadcast 4D to 4D',
'graph': {
'inputs': {
'inputA': {
'data': [6.2216410636901855],
'descriptor': {'dimensions': [1, 1, 1, 1], 'dataType': 'float32'}
},
'inputB': {
'data': [
-5.394711494445801, -7.189248561859131, -3.1081764698028564,
4.977657318115234, 5.111654281616211, -1.5386580228805542,
1.414366364479065, -0.9362112283706665, -6.029961585998535,
-3.0134198665618896, 0.170855313539505, 7.395327091217041,
7.178691864013672, -4.826237678527832, -2.020440101623535,
-3.267888069152832, 8.944384574890137, -5.932100772857666,
0.7069857120513916, 2.7764203548431396, 0.978833794593811,
-6.254901885986328, 4.409034729003906, -6.775286674499512
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'float32'}
}
},
'operators': [{
'name': 'greater',
'arguments': [{'a': 'inputA'}, {'b': 'inputB'}],
'outputs': 'output'
}],
'expectedOutputs': {
'output': {
'data': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1
],
'descriptor': {'dimensions': [2, 2, 2, 3], 'dataType': 'uint8'}
}
}
}
}
];
if (navigator.ml) {
greaterTests.forEach((test) => {
webnn_conformance_test(
buildGraphAndCompute, getGreaterPrecisionTolerance, test);
});
} else {
test(() => assert_implements(navigator.ml, 'missing navigator.ml'));
}
Codebase Browser
chromium