Skip to content

Fix embedding SDG dedup compatibility.#147

Open
hoangnguyenN94 wants to merge 1 commit intoNVIDIA-NeMo:mainfrom
hoangnguyenN94:hoannguyen/embedding-pineline
Open

Fix embedding SDG dedup compatibility.#147
hoangnguyenN94 wants to merge 1 commit intoNVIDIA-NeMo:mainfrom
hoangnguyenN94:hoannguyen/embedding-pineline

Conversation

@hoangnguyenN94
Copy link
Copy Markdown

Use Data Designer's public embedding facade API so Stage 0 dedup continues working after the private router interface change.

Use Data Designer's public embedding facade API so Stage 0 dedup continues working after the private router interface change.

Made-with: Cursor
# Data Designer's public embedding API moved behind ModelFacade methods.
# Use the facade directly rather than reaching into removed private attrs
# like `_router`, which breaks across library versions.
response = self.embedder.generate_text_embeddings(
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

@hoangnguyenN94 please could you update the data-designer dependency in the pyproject.toml of this retriever-sdg package to data-designer>=0.5.3 since that was the version that made the breaking change to _router

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

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

@oliverholworthy I have updated data-designer>=0.5.3, and re-build, but still error: 🛑 Failed to load plugin from entry point 'data-designer-retrieval-dedup': partially initialized module 'retriever_sdg.deduplication' has no attribute 'dd_retrieval_dedup_plugin' (most likely due to a circular import)

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