Why OpenAI API is very slow than ChatGPT

[kientv]

[kientv]

Why we need this chain, because it's only re-define question or translate question.
How to turn-off it to faster?

[AnasAli986]

How are you able to look at this screen? Is it langsmith? How do I set it up? Thanks

[kientv]

Yes, it's langsmith. You need setup langsmith env only. Maybe you need referral code to regist an account.

[HenryHengZJ]

Conversational Retrieval QA Chain block is seperated into 2 chains:

1. Reformat question using conversation history
2. Use the retrieved context to answer the reformatted question

You can take a look at the template - Prompt Chaining with VectorStore, its what executed under the hood

[kientv]

@HenryHengZJ Thank you.
I have just test "Prompt Chaining with VectorStore" without re-format question phrase.
It's much faster than. But LLMchain is not support Short Memory and "VectorStore to document" Search from vector store not working (every score).