Hi, I’m having some troubles with an LLM apps using Chroma for vector stores (via Langchain).
This the run time error message:
RuntimeError: �[91mYour system has an unsupported version of sqlite3. Chroma requires sqlite3 >= 3.35.0.�[0m�[94mPlease visit https://docs.trychroma.com/troubleshooting#sqlite to learn how to upgrade.�[0m
The strange thing is it has been working previously and still works in my local environment but not after deployed.
It seems that the version of SQLite on the deployment environment is not compatible with Chroma. The application seems to require SQLite version 3.35.0 or higher.
Does the version of SQLite in your deployed environment match that?
I deployed to Streamlit Cloud via Github - is there a way to configure the Streamlit container to meet the SQLite requirement for Chroma? It has been working fine before
@AnhNgDo According to this 🔍 Troubleshooting | Chroma you might try using a more recent version of python, which should come with a newer version of sqlite. Do you know what version you are using now on the app? You could try recreating the Cloud app with python 3.11 and see if that works.
I’m not sure which version of Python running on the deployed apps or how to check it. But that explains why it still works fine on my local machine (running 3.11).
How would I recreate the Cloud app with Python 3.11?
I have the some issue with this sqlite3 version. I tried Python 11 to create the app on Streamlit Cloud. It didn’t work. The sqlite3 version is still 3.34.1.
Actually, I tried Python 9, 10, 11 to create app on Streamlit cloud. The sqlite3 version is always 3.34.1.