Can we have opset >= 9 for DUC model?
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 9.8k
- Forks
- 1.6k
- PR merge metrics
- No merged PRs in 30d
Description
Model Request
Which model would you like to see in the model zoo?
object_detection_segmentation/duc
Describe why this model is relevant
Since current object_detection_segmentation/duc is opset 7, which keeps spatial as an attribute in operator:BatchNormalization.
But there is an issue: https://github.com/pytorch/pytorch/pull/9492.
In this PR, it only handle spatial=1, doesn't handle spatial != 1 condition.
In addition, current DUC model has BatchNormalization operator with spatial = 0. Therefore, DUC currently cannot be used.(Or my usage is not correct)
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the object_detection_segmentation/duc model in the model zoo and inspect its exported BatchNormalization operator, using the linked PyTorch issue for context. Compare the current opset 7 artifact with an opset of at least 9, and verify that the DUC model can be used without the spatial=0 incompatibility described here.
Written by the indexing model from the issue text.
Assessment
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100