Can we have langchain llm caching?

[OnlinePage]

Hi how can we implement it with llm caching either mongodb or sqlite , which serves uniquly to every user with caching to reduce api calls?

[msoedov]

Good question. There is a doc for caching https://python.langchain.com/docs/modules/model_io/models/llms/how_to/llm_caching.
You can add to prompt user_id to ensure that uniquely saved.

[OnlinePage]

can you give me a minimal example of it 🙂? Actually i know the langchain caching in general , but with langcorn to serve the caching with memory to every user uniquly is making me bit confused.
Pls a minimal example...

[OnlinePage]

@msoedov please can you provide an simple or minimal example for serving the cache with memory to every user uniquely 🥹