Find answers from the community

Updated last year

yes that one , but i want to understand

yes that one , but i want to understand something , whats the role of service_context in a llama index rag pipeline , when the tutorial set the gpt4 as service_contest.predictor, will we be using the custom embedding or open ai embedding when we are retreiving the document during the query decomposition ?

4 comments

TTeemu

It can be used for configuring resources like the embedding model and LLM being used, node parser etc.

TTeemu

By default it will use OpenAI embeddings

TTeemu

You need to define a different embedding model if you don't want to use OpenAI

TTeemu

Default LLM is also OpenAI (3.5) and by setting service context llm to gpt-4 you'll use that

Add a reply