adds retrieval logic

[koleshjr]

adds retrieval logic using LCEL
If any changes is required please let me know.
Thank you

[KevKibe]

Thanks for the PR.
Everything, checks out, don't mind the failed tests, I'll run the tests with my environment variables.
Is there a dependency that needs to be added to use
from operator import itemgetter and from langchain_core.prompts import PromptTemplate imports, if there is kindly add them to the PR in both requirements.txt and setup.cfg.
And could you remove one import of from langchain_pinecone import PineconeVectorStore module, it's imported twice in both files.

[koleshjr]

yeah only for langchain core though , I'm on my way home. will fix when I get home

[koleshjr]

Fixed it.
Added some flexibility to the retieve method by letting the user specify the models they want
But also adding default models incase they don't provide it
I have not tested them yet, Kindly test and lemme know if there are issues I need to fix

Thanks

[KevKibe]

LGTM

[KevKibe]

PR Merged🥳

[koleshjr]

🥳🥳🥳🥳

[koleshjr]

rerankers next. Still reading about them.

[koleshjr]

Btw kindly test the method. I just noticed a bug

retriever = vector_store.as_retriver(search_kwargs = {"k": top_k})

fix the above in both openai and google to :

retriever = vector_store.as_retriever(search_kwargs = {"k": top_k})

missed the e in the retriever

[KevKibe]

Nice.
I'm adding support for cohere embeddings and then do some tests to make sure everything works well, then I draft a release.

[koleshjr]

great