feat:AutoAlign guardrail updates

[KimiJL]

Description

API Changes: Alignment Configuration Name Change

• pii_fast -> pii
• confidential_detection -> confidential_info_detection
• text_toxicity_extraction -> toxicity_detection
• factcheck -> groundedness_checker

Added new config + flow for AutoAlign factchecker

A new fact check guardrail that takes the user message and bot message to generate a score from AutoAlign service, and depending on user set threshold from .co file (default to 0.5), a warning log will be produced with the score if below set threshold.
Previous factcheck/groundedness check has been fully renamed to groundedness check, and the same knowledge base factual scoring remains

Added multi_language parameter support

A new optional parameter that enables support for languages beyond English

Related Issue(s)

N/A

Checklist

• I've read the CONTRIBUTING guidelines.
• I've updated the documentation if applicable.
• I've added tests if applicable.
• @mentions of the person or team responsible for reviewing proposed changes.

[Pouyanpi]

Thank you @KimiJL, what do you think about adding a warning in the community section that there are backward incompatible changes in v0.12.0 version of the toolkit. As the flow names and config params have changed?

I'm going to take a closer look next week.

Please use

[PLACEHOLDER]{warning}
THE WARNING TEXT
[PLACEHOLDER]

change the [PLACEHOLDER] to ```

[KimiJL]

@Pouyanpi Thanks for the suggestion, it makes sense to me. I have added the warning under the updated guardrails, please let me know if that is the expected format.

[Pouyanpi]

Looks good to me, thanks 👍🏻