LlamaIndex
Log in

Find answers from the community

Home
Members
yasik
y
yasik
Offline, last seen 4 months ago
Joined September 25, 2024
yyasik
·

Hey team, a question about Phoenix (

Hey team, a question about Phoenix (arize) + OpenAIAssistantAgent + VectaraAutoRetriever. Currently in this setup it does not seem like Phoenix is collecting any internal steps that are executed by the VectaraAutoRetriever, so far we see only agent steps being collected which contain only the initial question and the final synthesized answer.

Do you have any tips on collecting intermediate steps by the VectaraAutoRetriever as well that would not require wrapping / rewriting of the retriever?
9 comments
y
L
yyasik
·

Looking at this model, it's a bit

Looking at this model, it's a bit unclear what's the difference between TOOL and FUNCTION. It appears TOOL is only used by openai_agent.py and step.py which is effectively the same module.

Perhaps it would be better to just use a single unified definition at the top model and wrap a special case (which is OpenAI). Or provide a more clear distinct naming, i.e. OPENAI_TOOL since it's a special case?

Plain Text
class MessageRole(str, Enum):
    """Message role."""

    SYSTEM = "system"
    USER = "user"
    ASSISTANT = "assistant"
    FUNCTION = "function"
    TOOL = "tool"
    CHATBOT = "chatbot"
4 comments
L
y
yyasik
·

hey team, I am trying to pair the

hey team, I am trying to pair the RetrieverQueryEngine that is configured with my auto retriever and REFINE response type as a QueryEngineTool and OpenAIAgent. While tracing I do see that the tool and auto retriever are invoked correctly, however, I don't see the query engine executing the refine steps. Am I missing something in this setup?
1 comment
L