Enhancement: Add ability to pass local cache to chat models

[eyurtsev]

Feature request discussed in #17176

Expand cache to accept a cache implementation in addition to a bool value:

langchain/libs/core/langchain_core/language_models/chat_models.py

Line 106 in 00a09e1

cache: Optional[bool] = None

If provided, will use the given cache.

Acceptance Criteria

• Documentation to the cache variable to explain how it can be used
• Update https://python.langchain.com/docs/modules/model_io/chat/chat_model_caching
• Include unit tests to test given functionality

PR can include implementation for caching of LLMs in addition to chat models.

[alhridoy]

Hi @eyurtsev Could you check my PR? Thanks!