Add support for all Granite Guardian risks

[martinscooper]

This PR adds support for all available Granite Guardian risks, extending current support of RAG risks, plus bug fixes and several improvements. In summary, this PR:

• Adds all supported risks (the ones available in this demo).
• Adds flexibility for input field names.
• Adds the concept of risks types: users risks, assistant risks, RAG risks, agentic risks and custom.
• Adds examples for every risk type.
• Adds catalog for every risk.
• Updates default model version to 3.1.
• Adds verification for early error detections.
• Adds debugging logs.
• Adds generation parameters to match those used in the Granite Guardian cookbook notebooks.
• Fix probability of risk score: certainty was being used.
• Adds support for 'Bring your own risk'.

[martinscooper]

I don't agree on calling 'question ' to the default user_message_field as it is not always a question. I would prefer user_message or user_prompt. This change may be not possible because of backward compatibility.

[martinscooper]

The RAG Granite Guardian metrics are duplicated now, but not exactly. The current Granite Guardian RAG metric is defined differently in the catalog. Copy is used to set data in the task data param (link to file), but there is no example on how to use it.

@elronbandel let me know if this is the way I should define the metrics in the prepare file and how should I adapt the example.

[martinscooper]

I also changed the Granite-based LLM evaluators. I removed the '0' from the version as it looked awkward. Let me know if it is to late to make that change and I will revert it.

[martinscooper]

@arielge FYI, we are adding all the missing Granite Guardian risks.

[elronbandel]

@martinscooper is this breaking change or just internal renaming?