microsoft / microsoft/onnxruntime

[ORT CPU][ORT GPU (DML)][WebNN] Fail to run WebNN batchNormalization tests with 1D tensor due to "Dimension mismatch in unification between 6 and 1"

Open
#26,521 0 comments 0 reactions 1 assignee View on GitHub

@hariharans29 is already working on this.

Since Nov 12, 2025.

ep:DML ep:WebNN WebNN
Dominant language
C++
Stars
21.9k
Forks
4.2k
Avg merge
4d 11h
Merged PRs (30d)
184

Description

### Describe the issue

Fail to run WebNN batchNormalization tests with 1D tensor by ORT default CPU EP and ORT default GPU DML EP due to below error

>error message: Node (_0) Op (BatchNormalization) [ShapeInferenceError] Dimension mismatch in unification between 6 and 1

/cc @fdwr PTAL, thanks!

### To reproduce

1. Install Windows App SDK Stable [1.8.2 (1.8.251003001)](https://learn.microsoft.com/en-us/windows/apps/windows-app-sdk/stable-channel#version-18)
2. Install [lates Edge Canary browser](https://www.microsoft.com/en-us/edge/download/insider?form=MA13FJ)
3. Launch Edge Canary browser, navigate to `about://flags` enabled "Enables WebNN API" and "ONNX Runtime backend for WebNN" flags, then relaunch browser
4. Navigate to this test url `https://wpt.live/webnn/conformance_tests/batch_normalization.https.any.html?device=cpu&tc=batchNormalization%20float32%201D%20tensor%20options.axis=0` by testing ORT default CPU EP or `https://wpt.live/webnn/conformance_tests/batch_normalization.https.any.html?device=gpu&tc=batchNormalization%20float32%201D%20tensor%20options.axis=0` by testing ORT default GPU EP
5. We can find "batchNormalization float32 1D tensor options.axis=0" test run Fail with the error of "UnknownError: Failed to execute 'build' on 'MLGraphBuilder': Failed to create session."
6. Navigate to `about://gpu`, we can get the log message like
>error message: Node (_0) Op (BatchNormalization) [ShapeInferenceError] Dimension mismatch in unification between 6 and 1

Same issue to run `batchNormalization float16 1D tensor options.axis=0` of fp16 test, test url `https://wpt.live/webnn/conformance_tests/batch_normalization.https.any.html?device=cpu&tc=batchNormalization%20float16%201D%20tensor%20options.axis=0` by testing ORT default CPU EP or `https://wpt.live/webnn/conformance_tests/batch_normalization.https.any.html?device=gpu&tc=batchNormalization%20float16%201D%20tensor%20options.axis=0` by testing ORT default GPU EP
### Urgency

_No response_

### ONNX Runtime Installation

Released Package Windows App SDK Stable [1.8.2 (1.8.251003001)](https://learn.microsoft.com/en-us/windows/apps/windows-app-sdk/stable-channel#version-18)

### ONNX Runtime Version or Commit ID

1.23.25.928

### Execution Provider

ORT default CPU EP, ORT default GPU EP

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.