Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ModelCardGenerator component cannot run in Vertex AI #259

Open
jeongukjae opened this issue Jul 25, 2023 · 4 comments · May be fixed by #260
Open

ModelCardGenerator component cannot run in Vertex AI #259

jeongukjae opened this issue Jul 25, 2023 · 4 comments · May be fixed by #260

Comments

@jeongukjae
Copy link
Contributor

Expected Behavior

Actual Behavior

Vertex AI requires artifact schema title to be ^[a-z][a-z0-9-_]{2,20}[.][A-Z][a-zA-Z0-9-_]{2,49}$, and model card generator component has output artifact typed tfx_addons.model_card_generator.artifact.ModelCard. So it cannot be run in Vertex AI.

Error message:

com.google.cloud.ai.platform.common.errors.AiPlatformException: code=INVALID_ARGUMENT, message=MetadataSchema must include a schema title that satisfies the following regex: ^[a-z][a-z0-9-_]{2,20}[.][A-Z][a-zA-Z0-9-_]{2,49}$, cause=null

Steps to Reproduce the Problem

Specifications

tensorflow==2.9.3
tfx==1.9.1
tfx-addons==0.6.0
...
  • Version:
  • Platform: Vertex AI

Related issue:

@jeongukjae
Copy link
Contributor Author

Oh, it was resolved at v1.12. Closing this issue.

@jeongukjae
Copy link
Contributor Author

Oh, not resolved in v1.12! sorry.

@jeongukjae jeongukjae reopened this Jul 25, 2023
@jeongukjae
Copy link
Contributor Author

In v1.12, there are another errors.

ModelCard artifact does not satisfy the schema title regex (^[a-z][a-z0-9-_]{2,20}[.][A-Z][a-zA-Z0-9-_]{2,49}$, no nested classpath).
So ModelCard artifact's TYPE_NAME should be updated.

https://github.com/tensorflow/tfx/blob/2fd4f95ce3e631e7763611fd8ed631d59bc053d8/tfx/orchestration/kubeflow/v2/compiler_utils.py#L302-L335

ModelCard artifact type should be accessible from tfx_addons (top-level module) to be resolved in kubeflow v2's container entrypoint.
So tfx-addons should export all artifact types from top-level.

https://github.com/tensorflow/tfx/blob/2fd4f95ce3e631e7763611fd8ed631d59bc053d8/tfx/orchestration/kubeflow/v2/container/kubeflow_v2_entrypoint_utils.py#L219-L234

@jeongukjae jeongukjae linked a pull request Jul 25, 2023 that will close this issue
2 tasks
@jeongukjae
Copy link
Contributor Author

Same issue with #249

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 a pull request may close this issue.

1 participant