Skip to content

Conversation

TomWildenhain-Microsoft
Copy link
Collaborator

Signed-off-by: Tom Wildenhain [email protected]

dtypes = rename_tensors_in_dict(dtypes)
inputs_as_nchw = rename_tensors_in_list(inputs_as_nchw)
if not is_subgraph:
rename_tensors_in_nodes(onnx_nodes)
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We actually don't need to rename tensors in subgraphs at all.

@TomWildenhain-Microsoft TomWildenhain-Microsoft enabled auto-merge (squash) April 27, 2021 19:50
Copy link
Contributor

@guschmue guschmue left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

interesting that we did not notice this.
maybe we need to find a nice tf-2 model for integration tests.

@TomWildenhain-Microsoft TomWildenhain-Microsoft merged commit b002761 into master Apr 27, 2021
@TomWildenhain-Microsoft TomWildenhain-Microsoft deleted the tom/fix_subgraph_bug branch April 27, 2021 21:12
zerollzeng pushed a commit to zerollzeng/tensorflow-onnx that referenced this pull request May 16, 2021
* Fix bug that renamed subgraph i/o twice

Signed-off-by: Tom Wildenhain <[email protected]>

* Don't rename tensors in subgraphs at all

Signed-off-by: Tom Wildenhain <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants