[WebNN] Support Cast fusion specific for int64 data type #23256
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Some WebNN backends do not support the int64 data type, but this limitation can be addressed by converting the model's int64 inputs, outputs, and initializers to int32. However, certain ONNX nodes, such as ArgMax, ArgMin, and ScatterND, require the int64 data type for specific inputs or outputs. To handle such case, we can add Cast nodes before or after these nodes in the model and fuse them during WebNN EP optimization. The fusion strategy is as follows: 1. Verify if the Cast node can be fused with either the preceding node or the successive node. 2. Check if the node requiring the int64 data type can be supported solely by addressing the int64 data type limitation. Ensure that the node is unsupported only due to the int64 restriction. 3. Use an is_fusable flag to record paired nodes as <Cast node index, fusable node index> that can be fused together. 4. Mark the fusable nodes as supported after identifying them. 5. During WebNN graph compilation, skip the Cast node and fuse it in its paired fusable node.
|
/azp run ONNX Runtime Web CI Pipeline,Windows GPU CI Pipeline,Linux Android Emulator QNN CI Pipeline |
|
/azp run Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline,Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,Windows ARM64 QNN CI Pipeline,Windows CPU CI Pipeline |
|
/azp run Windows GPU TensorRT CI Pipeline,onnxruntime-binary-size-checks-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,Windows x64 QNN CI Pipeline,Big Models |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
/azp run Windows GPU CUDA CI Pipeline,Windows GPU DML CI Pipeline,Windows GPU Doc Gen CI Pipeline |
|
Azure Pipelines successfully started running 4 pipeline(s). |
|
Azure Pipelines successfully started running 3 pipeline(s). |
|
Azure Pipelines successfully started running 9 pipeline(s). |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Some WebNN backends do not support the int64 data type, but this limitation can be addressed by converting the
model's int64 inputs, outputs, and initializers to int32. However, certain ONNX nodes, such as ArgMax, ArgMin,
and ScatterND, require the int64 data type for specific inputs or outputs.
To handle such case, we can add Cast nodes before or after these nodes in the model and fuse them during WebNN EP
optimization. The fusion strategy is as follows:
limitation. Ensure that the node is unsupported only due to the int64 restriction.