feat: enabled saving and evaluation for moderator

[JXZhou0224]

Closes #

📑 Description

added two new features for moderator:

saving

1. Now LLMAgents will store their profile in a AgentProfile on redis.
2. Later, LLMAgent can be loaded from existing AgentProfiles
3. and moderator can access AgentProfile to generate an EpisodeLog

evaluation

1. created evaluators for multiagent assessment (current Evaluators in sotopia only supports two agents)
2. moderator can now evaluate an episode on demand, currently there is only one evaluator available, new evaluators will be added later.

✅ Checks

• My pull request adheres to the code style of this project
• My code requires changes to the documentation
• I have updated the documentation as required
• All the tests have passed
• Branch name follows type/descript (e.g. feature/add-llm-agents)
• Ready for code review

ℹ Additional Information

[XuhuiZhou]

why not using the existing evaluators

[JXZhou0224]

The existing evaluators base class can only take in message history. However, there might be cases where we need other information in EpisodeLog to generate an evaluation. So I made a new class of evaluator for future use.