This article shows how to quickly build chat applications using Python and leveraging powerful technologies such as OpenAI ChatGPT models, Embedding models, LangChain framework, ChromaDB vector datab...
Updated May 07, 2024
Version 8.0
Blog Post
55 MIN READ
Create an Azure OpenAI, LangChain, ChromaDB, and Chainlit Chat App in Container Apps using Terraform
Thanks Heman_k
I'm not an Azure Cognitive Search subject-matter expert, but I can surely affirm that Azure Cognitive Search has strong vector search capabilities that allow you to search and retrieve similar vectors based on their similarity scores. While I do not have direct comparisons to other similar products or other vector databases, such as Chroma or FAISS, Azure Cognitive Search's vector search capabilities are designed to provide efficient and accurate results.
Regarding the HNSW algorithm, Azure Search does support it. The HNSW algorithm is a commonly used algorithm for approximate nearest neighbor search, and it is available in Azure Cognitive Search for vector search scenarios. Please check Add vector search - Azure Cognitive Search | Microsoft Learn.
As for the specific issue you mentioned regarding importing the HnswVectorSearchAlgorithmConfiguration module in Python 3.9, it's possible that the module may not be available or accessible in that specific version. It is recommended to check the documentation or reach out to Microsoft Azure support for further assistance or alternative approaches.
Overall, Azure Search offers robust vector search capabilities, including support for the HNSW algorithm, which can be valuable for creating efficient and accurate search experiences.
P.S. If you found my article and sample interesting and helpful, please like the article and star the GitHub project, thanks 🙂