Find answers from the community

Updated 3 weeks ago

the issue with vector index auto retriever

Hello, I am trying to build VectorIndexAutoRetriver and it is working until the filters applied.

if the filter is applied it is giving noneType error. I am adding metadata but I don't know what is the issue. I am new so please help me
R
L
W
8 comments
2025-01-24 17:18:48,301 - llama_index.core.indices.vector_store.retrievers.auto_retriever.auto_retriever - INFO - Using query str: summary of latest records
2025-01-24 17:18:48,301 - llama_index.core.indices.vector_store.retrievers.auto_retriever.auto_retriever - INFO - Using filters: []
Using query str: summary of latest records
Using filters: []
2025-01-24 17:18:48,304 - llama_index.core.indices.vector_store.retrievers.auto_retriever.auto_retriever - INFO - Using top_k: 4
Retrieval entering 8ea6a13f-d522-4821-8e7a-8717ecf0ceb6: VectorIndexRetriever
Retrieving from object VectorIndexRetriever with query summary of latest records
2025-01-24 17:19:01,034 - root - ERROR - argument of type 'NoneType' is not iterable
there should be an actual traceback no?
Hard to say what the issue is without a traceback
no, there is not. that is why I am confuse as what is the issue?
Theres always a traceback. Just depends on where/how you are running your code

Factor out your code into a script that reproduce. Sounds like you are maybe running in a fastapi server?
yes i am using fastapi
Reproduce it outside of fastapi then maybe?
Did you try with traceback module as well?
Add a reply
Sign up and join the conversation on Discord