mattgotteiner
commented
3 months ago Hi @sureshmoon ,
We just updated langchain to enable you to specify this property. Once a new version of langchain is published, we'll update the sample to use this version so you can specify connection_verify when creating the azure search vector store
Hi,
I am experiencing an SSL verification error while working with the sample_rag_langchain.ipynb notebook in the Python folder of the Document Intelligence Code Samples repository. Specifically, this occurs in the section where embeddings and indexes are created using Azure Search.
Error Details When running the cell that uses the AzureSearch class, the following error is raised: ServiceRequestError: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1000)
Repository: Azure Samples Document Intelligence Code Samples Folder:Python(v4.0)/Retrieval_AugmentedGeneration(RAG)_samples) Notebook: sample_rag_langchain.ipynb Platform: Running within a company portal, windows 11.
Issue In the native Azure Search library, I typically resolve this issue by setting the parameter connection_verify=False. However, the langchain library does not seem to have an option to disable SSL verification.
Request Could you please provide guidance or an update to the notebook/library to include an option to disable SSL verification for AzureSearch in the langchain library? This would help in bypassing the SSL certificate verification error.
Thank you for your assistance.