[feat] Support Amazon Bedrock

[moshe-shelly]

Description
Add support for Amazon Bedrock in the Guardrails package to expand its compatibility with large language models (LLMs) and AI services. This feature would allow users to integrate Guardrails' validation and safety mechanisms with Amazon's foundation models.

Why is this needed

• Expanding Ecosystem: Amazon Bedrock is a significant player in the AI services market, and supporting it would broaden Guardrails' applicability.
• Enterprise Adoption: Many enterprises use AWS services, including Bedrock. This integration would make Guardrails more attractive to these organizations.
• Model Diversity: Bedrock offers access to various foundation models, including those from Anthropic, AI21 Labs, and Amazon's own models. This diversity can enhance the capabilities of applications using Guardrails.
• Seamless AWS Integration: For users already leveraging AWS services, having Guardrails support Bedrock would allow for a more cohesive and streamlined development experience.
• Enhanced Security: Combining Guardrails' safety features with Bedrock's enterprise-grade security could provide a robust solution for organizations with strict compliance requirements.

End result
Same as the native support for Anthropic, OpenAI, Cohere etc.

[CalebCourier]

Amazon Bedrock is already supported through our LiteLLM integration (see LiteLLM docs for details on usage here). In v0.6.0 we dropped native support for LLM's including Anthropic, OpenAI, and Cohere in favor of standardized aggregators like LiteLLM and Manifest.

@zsimjee any other thoughts or action items you want to take before closing?

[zsimjee]

Nope, please refer to the litellm documentation to see which env vars to set

https://docs.litellm.ai/docs/providers/bedrock.

I also added info here for the next docs release
#1207