non-working input standardization

[zsimjee]

This does not currently work.

This pr attempts to use input/output mapping to deal with LLMs in a more generic way. What I've discovered is that there's simply too much going on with both our existing implementation as well as in individual LLM clients. I think that overhauling to this level is too big a change for the minor version update with regards to the risk and testing surface area involved.

This PR is missing a few key parts:

1. mapping inputs for the injection of schema and reask instructions
2. dealing with function calling
3. testing around streaming, async

Fundamentally, I do think this approach can work, and is a better mapping solution than the one that we currently have since in some ways it's "looser". It does not include recreating clients or trying to type-match into them, which I think is a huge benefit from a maintainability perspective.

All that being said, I would propose a difference in deliverables for 0.4.3 that I think are still achievable, help with debug scenarios, and keep backwards compatability without being too huge a change. However, I only think these changes are appropriate if we do necessary pruning on 0.5.0.

0.4.3

1. make the __call__ llm_api callable optional. When it is optional, we pass through information to a LiteLLM client we create internally
2. expose a naked_llm_call function on guard. This function would have inputs styled the same way we style __call__ inputs, but purely uses our internal mapping to make a call to the LLM without guarding. This would help users debug and translate from their call pattern to the guard-style call pattern.

0.5.0

1. delete all custom mappings in llm_providers other than OpenAICallable, LiteLLMCallable, and Sync/AsycnBaseCallable
2. make clear in docs that those callables are the only supported ones. If you have a callable other than those and don't want to wrap in a BaseCallable, then orchestrate your own llm api callable and
3. remove msg_history, prompt, instructions. Expect messages for BaseCallable. For everything else, pass through args as is
4. in OpenAI, cohere, and anthropic, never recreate a client
5. get rid of old openai versions across the codebase
6. create 1 doc notebook with examples for each llm type present in llms.py
7. support streaming and asycn for openai and litellm callables natively but through a similar output mapping strategy as found in this PR.