have a question for all of you. Feel free to reply, and if you can respond quickly, that would be great. When we use a vector store index in the backend, what is used to create the embeddings? Also, if we don't mention our API key, will it create embeddings, or where do we specify the API key in the vector store index? Can someone clarify how this works?
We have to provide an API key, otherwise it won't work, because this model (OpenAI's text-embedding-ada-002) is not open source. Where do we mention the API key for this model?
One more thing: what is the main job of a vector store index? It's used to create embeddings with third-party models, so why are there different kinds of indexes, and what are their use cases? If we're using a third-party model to create embeddings and storing them somewhere else, what is the purpose of the vector store index, and why are there different types of indexes?
Can anyone help explain why LlamaIndex is better for indexing and retrieval compared to the LangChain framework? What makes it different from LangChain? If someone can answer this properly, there's a reward, this question asked by me in interview .
