docs: update RagasEvaluator to reflect Ragas 0.4.3 API changes

[ritikraj2425]

Related Issues

• fixes docs: Update Ragas docs #11178

Proposed Changes:

Update the RagasEvaluator documentation to reflect the breaking changes in the Ragas 0.4.3 integration:

• Modernized Metric Initialization: Switched from the RagasMetric enum and metric_params to the modern Ragas metrics API. Documentation now reflects that metrics should be initialized as classes from ragas.metrics.collections (e.g., AnswerRelevancy(llm=llm)).
• Updated run() Method: Reflected the change in the run() method signature. The component now accepts explicit arguments like query, response, documents, and reference instead of a nested inputs dictionary.
• LLM Configuration: Updated code examples to use AsyncOpenAI and llm_factory, which is the new recommended pattern for configuring Ragas metrics.
• Cleanup: Removed a duplicated parameter table in the ragasevaluator.mdx file.

How did you test it?

• Manual verification of code snippets against the new RagasEvaluator implementation in the haystack-core-integrations repository (PR #3207).
• Verified that the example pipeline structure correctly maps to the new component input/output requirements.

Notes for the reviewer

• These changes bring the manual .mdx documentation in sync with the automatically generated API reference (which was already updated in the main repository).

Checklist

• I have read the contributors guidelines and the code of conduct.
• I have updated the related issue with new insights and changes.
• I have added unit tests and updated the docstrings. (N/A - Documentation only)
• I've used one of the conventional commit types for my PR title: docs:
• I have documented my code.
• I have added a release note file, following the contributors guidelines. (N/A for documentation-only changes)
• I have run pre-commit hooks and fixed any issue.

[vercel]

@ritikraj2425 is attempting to deploy a commit to the deepset Team on Vercel.

A member of the Team first needs to authorize it.

[sjrl]

Hey @ritikraj2425 thanks for opening! Could you handle the merge conflict?

[sjrl]

Let's add a note that the env variable OPENAI_API_KEY must be set for the code block below to work.

[sjrl]

If we want to make this an example about AnswerRelevancy we also need to set the embeddings param

embeddings=embedding_factory("openai", model="text-embedding-3-small", client=client)
AnswerRelevancy(llm=llm, embeddings=embeddings)

[sjrl]

Please run this code block locally to make sure it works