commit 72be82731cd456a588e95c99b319f7ee76710710
parent 978f01a16fcfd70d84c6547f4fafbdb6f3962f0f
Author: Phillis Tang <phillis@chromium.org>
Date: Wed, 3 Dec 2025 14:49:55 +0000
Bug 2003749 [wpt PR 56433] - webnn: move constant tests out of batch_normalization, a=testonly
Automatic update from web-platform-tests
webnn: move constant tests out of batch_normalization
The constant tests are failing on Linux GPU, and for fp16 it's failing
flakily because of precision. We can't disable flaky tests
individually for a WPT file. So seperate out the failing tests to
a new file.
Bug: 460534989
Change-Id: Ic391cc67b058f6ffad9193684d7912aa144245dc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7204448
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Commit-Queue: Phillis Tang <phillis@chromium.org>
Reviewed-by: David Baron <dbaron@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1553060}
--
wpt-commits: 68bce81c6240f79769819649ed9fc1169b474bf9
wpt-pr: 56433
Diffstat:
2 files changed, 135 insertions(+), 106 deletions(-)
diff --git a/testing/web-platform/tests/webnn/conformance_tests/batch_normalization.https.any.js b/testing/web-platform/tests/webnn/conformance_tests/batch_normalization.https.any.js
@@ -128,65 +128,6 @@ const batchNormTests = [
}
},
{
- 'name': 'batchNormalization float32 2D constant tensors default options',
- 'graph': {
- 'inputs': {
- 'bnInput': {
- 'data': [
- -41.30733108520508, 64.08863830566406, -63.376670837402344,
- -46.790367126464844, 83.02227020263672, -80.08049011230469,
- -62.144378662109375, -0.10012771934270859, -40.90216064453125,
- 56.96306228637695, 37.37249755859375, 57.046478271484375,
- 82.05680084228516, -86.1164321899414, 76.8831787109375,
- 97.03362274169922, -21.35103988647461, -96.93824005126953,
- -9.359310150146484, 80.20824432373047, -85.36802673339844,
- 62.35185241699219, -68.4724349975586, -12.10716724395752
- ],
- 'descriptor': {shape: [4, 6], dataType: 'float32'},
- 'constant': true
- },
- 'bnMean': {
- 'data': [
- -7.814267635345459, -95.64129638671875, 38.15440368652344,
- -55.95203399658203, -87.86500549316406, -41.63645553588867
- ],
- 'descriptor': {shape: [6], dataType: 'float32'},
- 'constant': true
- },
- 'bnVariance': {
- 'data': [
- 60.31186294555664, 26.43260383605957, 53.275634765625,
- 40.146121978759766, 59.41098403930664, 35.99981689453125
- ],
- 'descriptor': {shape: [6], dataType: 'float32'},
- 'constant': true
- }
- },
- 'operators': [{
- 'name': 'batchNormalization',
- 'arguments': [
- {'input': 'bnInput'}, {'mean': 'bnMean'}, {'variance': 'bnVariance'}
- ],
- 'outputs': 'bnOutput'
- }],
- 'expectedOutputs': {
- 'bnOutput': {
- 'data': [
- -4.312741756439209, 31.068212509155273, -13.910240173339844,
- 1.4459478855133057, 22.170541763305664, -6.407354354858398,
- -6.995829105377197, 18.583200454711914, -10.831125259399414,
- 17.820920944213867, 16.2480411529541, 16.447195053100586,
- 11.57226848602295, 1.8526301383972168, 5.306026458740234,
- 24.145092010498047, 8.629376411437988, -9.216986656188965,
- -0.1989477425813675, 34.203548431396484, -16.923160552978516,
- 18.671411514282227, 2.5159497261047363, 4.921559810638428
- ],
- 'descriptor': {shape: [4, 6], dataType: 'float32'}
- }
- }
- }
- },
- {
'name': 'batchNormalization float32 2D tensor default options',
'graph': {
'inputs': {
@@ -828,53 +769,6 @@ const batchNormTests = [
}
},
{
- 'name': 'batchNormalization float16 2D constant tensors default options',
- 'graph': {
- 'inputs': {
- 'bnInput': {
- 'data': [
- -41.3125, 64.0625, -63.375, -46.78125, 83,
- -80.0625, -62.15625, -0.10009765625, -40.90625, 56.96875,
- 37.375, 57.03125, 82.0625, -86.125, 76.875,
- 97.0625, -21.34375, -96.9375, -9.359375, 80.1875,
- -85.375, 62.34375, -68.5, -12.109375
- ],
- 'descriptor': {shape: [4, 6], dataType: 'float16'},
- 'constant': true
- },
- 'bnMean': {
- 'data': [-7.8125, -95.625, 38.15625, -55.9375, -87.875, -41.625],
- 'descriptor': {shape: [6], dataType: 'float16'},
- 'constant': true
- },
- 'bnVariance': {
- 'data': [60.3125, 26.4375, 53.28125, 40.15625, 59.40625, 36],
- 'descriptor': {shape: [6], dataType: 'float16'},
- 'constant': true
- }
- },
- 'operators': [{
- 'name': 'batchNormalization',
- 'arguments': [
- {'input': 'bnInput'}, {'mean': 'bnMean'}, {'variance': 'bnVariance'}
- ],
- 'outputs': 'bnOutput'
- }],
- 'expectedOutputs': {
- 'bnOutput': {
- 'data': [
- -4.3125, 31.0625, -13.90625, 1.4453125, 22.171875,
- -6.40625, -6.99609375, 18.578125, -10.828125, 17.8125,
- 16.25, 16.4375, 11.5703125, 1.84765625, 5.3046875,
- 24.140625, 8.6328125, -9.21875, -0.19921875, 34.1875,
- -16.921875, 18.671875, 2.513671875, 4.91796875
- ],
- 'descriptor': {shape: [4, 6], dataType: 'float16'}
- }
- }
- }
- },
- {
'name': 'batchNormalization float16 2D tensor default options',
'graph': {
'inputs': {
diff --git a/testing/web-platform/tests/webnn/conformance_tests/batch_normalization_constant.https.any.js b/testing/web-platform/tests/webnn/conformance_tests/batch_normalization_constant.https.any.js
@@ -0,0 +1,135 @@
+// META: title=test WebNN API batchNormalization operation with constant inputs
+// META: global=window
+// 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-batchnorm
+// Normalize the values of the input tensor using Batch-Normalization.
+//
+// dictionary MLBatchNormalizationOptions {
+// MLOperand scale;
+// MLOperand bias;
+// [EnforceRange] unsigned long axis = 1;
+// double epsilon = 1e-5;
+// };
+//
+// MLOperand batchNormalization(
+// MLOperand input, MLOperand mean, MLOperand, variance,
+// optional MLBatchNormalizationOptions options = {});
+
+const batchNormTests = [
+ {
+ 'name': 'batchNormalization float32 2D constant tensors default options',
+ 'graph': {
+ 'inputs': {
+ 'bnInput': {
+ 'data': [
+ -41.30733108520508, 64.08863830566406, -63.376670837402344,
+ -46.790367126464844, 83.02227020263672, -80.08049011230469,
+ -62.144378662109375, -0.10012771934270859, -40.90216064453125,
+ 56.96306228637695, 37.37249755859375, 57.046478271484375,
+ 82.05680084228516, -86.1164321899414, 76.8831787109375,
+ 97.03362274169922, -21.35103988647461, -96.93824005126953,
+ -9.359310150146484, 80.20824432373047, -85.36802673339844,
+ 62.35185241699219, -68.4724349975586, -12.10716724395752
+ ],
+ 'descriptor': {shape: [4, 6], dataType: 'float32'},
+ 'constant': true
+ },
+ 'bnMean': {
+ 'data': [
+ -7.814267635345459, -95.64129638671875, 38.15440368652344,
+ -55.95203399658203, -87.86500549316406, -41.63645553588867
+ ],
+ 'descriptor': {shape: [6], dataType: 'float32'},
+ 'constant': true
+ },
+ 'bnVariance': {
+ 'data': [
+ 60.31186294555664, 26.43260383605957, 53.275634765625,
+ 40.146121978759766, 59.41098403930664, 35.99981689453125
+ ],
+ 'descriptor': {shape: [6], dataType: 'float32'},
+ 'constant': true
+ }
+ },
+ 'operators': [{
+ 'name': 'batchNormalization',
+ 'arguments': [
+ {'input': 'bnInput'}, {'mean': 'bnMean'}, {'variance': 'bnVariance'}
+ ],
+ 'outputs': 'bnOutput'
+ }],
+ 'expectedOutputs': {
+ 'bnOutput': {
+ 'data': [
+ -4.312741756439209, 31.068212509155273, -13.910240173339844,
+ 1.4459478855133057, 22.170541763305664, -6.407354354858398,
+ -6.995829105377197, 18.583200454711914, -10.831125259399414,
+ 17.820920944213867, 16.2480411529541, 16.447195053100586,
+ 11.57226848602295, 1.8526301383972168, 5.306026458740234,
+ 24.145092010498047, 8.629376411437988, -9.216986656188965,
+ -0.1989477425813675, 34.203548431396484, -16.923160552978516,
+ 18.671411514282227, 2.5159497261047363, 4.921559810638428
+ ],
+ 'descriptor': {shape: [4, 6], dataType: 'float32'}
+ }
+ }
+ }
+ },
+ {
+ 'name': 'batchNormalization float16 2D constant tensors default options',
+ 'graph': {
+ 'inputs': {
+ 'bnInput': {
+ 'data': [
+ -41.3125, 64.0625, -63.375, -46.78125, 83,
+ -80.0625, -62.15625, -0.10009765625, -40.90625, 56.96875,
+ 37.375, 57.03125, 82.0625, -86.125, 76.875,
+ 97.0625, -21.34375, -96.9375, -9.359375, 80.1875,
+ -85.375, 62.34375, -68.5, -12.109375
+ ],
+ 'descriptor': {shape: [4, 6], dataType: 'float16'},
+ 'constant': true
+ },
+ 'bnMean': {
+ 'data': [-7.8125, -95.625, 38.15625, -55.9375, -87.875, -41.625],
+ 'descriptor': {shape: [6], dataType: 'float16'},
+ 'constant': true
+ },
+ 'bnVariance': {
+ 'data': [60.3125, 26.4375, 53.28125, 40.15625, 59.40625, 36],
+ 'descriptor': {shape: [6], dataType: 'float16'},
+ 'constant': true
+ }
+ },
+ 'operators': [{
+ 'name': 'batchNormalization',
+ 'arguments': [
+ {'input': 'bnInput'}, {'mean': 'bnMean'}, {'variance': 'bnVariance'}
+ ],
+ 'outputs': 'bnOutput'
+ }],
+ 'expectedOutputs': {
+ 'bnOutput': {
+ 'data': [
+ -4.3125, 31.0625, -13.90625, 1.4453125, 22.171875,
+ -6.40625, -6.99609375, 18.578125, -10.828125, 17.8125,
+ 16.25, 16.4375, 11.5703125, 1.84765625, 5.3046875,
+ 24.140625, 8.6328125, -9.21875, -0.19921875, 34.1875,
+ -16.921875, 18.671875, 2.513671875, 4.91796875
+ ],
+ 'descriptor': {shape: [4, 6], dataType: 'float16'}
+ }
+ }
+ }
+ }
+];
+
+webnn_conformance_test(
+ batchNormTests, buildAndExecuteGraph, getPrecisionTolerance);
