How to use LLMCaching

[brooda]

Hi,

I wanted to use Langchain LLMCaching - https://python.langchain.com/docs/how_to/llm_caching/
How I can use it with trustcall extractor (when input is the same - I would just like to use the cache)?

Just using
set_llm_cache(SQLiteCache(database_path=".langchain.db"))
does not work

[hinthornw]

Hm it should work as-is, so long as the langchain caching mechanism is working correctly - so you're not getting any matches?

[jasonbhart]

Confirming it should just work out of the box, as long as you setup caching on the LLM before you create a trustcall extractor with it. Its working here. I'd suggest confirming caching works as expected before introducing trustcall to the mix.