[TorchOnnxToTorch] Fix DepthToSpace conversion in CRD mode #4351
+30
−16
Conversation
This file contains hidden or 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
According to the ONNX spec, in CRD mode the input should be reshaped to [b, c / (blocksize^2), blocksize, blocksize, h, w] instead of [b, blocksize, blocksize, c / (blocksize^2), h, w].
|
@sahas3 Could you please review this PR and share feedback or suggestions when you have time? Thanks a lot! |
sahas3
requested changes
Oct 29, 2025
| Value transposedInput; | ||
| Value reshapeSizesList; | ||
| if (mode == "DCR") { | ||
| reshapeSizesList = rewriter.create<Torch::PrimListConstructOp>( |
There was a problem hiding this comment.
I am not familiar with the ONNX part of the codebase. Are there e2e tests that will lock down the numeric for DCR/CRD mode? If not, can you add some please?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
According to the ONNX spec, in the CRD mode, elements along the depth dimension from the input tensor are rearranged in the following order: column, row, and the depth. The output y is computed from the input x as below:
So, the input should be reshaped to [b, c / (blocksize^2), blocksize, blocksize, h, w] instead of [b, blocksize, blocksize, c / (blocksize^2), h, w].