Reranking is a technique that indexes the documents from most to least semantically relevant to the query. 🌟LanceDB now supports Jina AI reranker out of the box. This example walkthrough takes you through improving retrieval performance on a synthetic dataset from a baseline

🌟 "Every Continue user is a LanceDB user!" says Nate Sesti Co-founder and CTO of Continue At LanceDB, we're proud to empower businesses like Continue with our powerful vector database solution. 🚀 Join our Discord channel discord.gg/G5DcmnZWKB

🔒 Privacy meets performance with LanceDB and AnythingLLM "Of all instances of AnythingLLM, 96% use our default, and recommended, vector database, LanceDB for storage and retrieval of their documents. LanceDB allows all users the best of both performance for vector search as

🚀 New Tutorial for AI Engineers! 🚀 We have just published a comprehensive guide on integrating Twelve Labs' Embed API with LanceDB to build advanced video understanding applications 👇

💻 𝗧𝗵𝗲 𝗻𝗲𝘄 𝗟𝗮𝗻𝗰𝗲 𝗳𝗼𝗿𝗺𝗮𝘁 𝘃𝟮 𝗶𝘀 𝗿𝗲𝗮𝗱𝘆 𝗶𝗻 𝗯𝗲𝘁𝗮! blog.lancedb.com/lance-v2-is-no… We've been talking for a while about a new iteration of our file format. We're pleased to announce that the new v2 format is now in available as an opt-in feature for lance and

Published a report on improving retrievers using LanceDB Hybrid search and Reranking. This explores techniques to optimize retrieval performance 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗿𝗲𝗾𝘂𝗶𝗿𝗶𝗻𝗴 𝗮 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗱𝗮𝘁𝗮𝘀𝗲𝘁 𝗿𝗲-𝗶𝗻𝗴𝗲𝘀𝘁𝗶𝗼𝗻 blog.lancedb.com/hybrid-search-…

Lance v0.16.1 brings new features to help manage versions and Lance file format editions, a new optimized ETL pattern for updating columns, and an experimental API for distributed ANN index training. Read more at: blog.lancedb.com/lance-v0-16-1-…

In the last few of weeks we've made many docs & feature improvements at LanceDB based on user feedback. Here are a few: 1. More flexible hybrid search API - Users love lancedb hybrid search with embedding API. Now, you can also use it without automatic vectorisation.

Studying about variational generative models is never complete without writing VAEs from scratch. Here is how I did it with the lance data format from LanceDB blog.lancedb.com/training-a-var…

Blog Alert: From blog to 🤖: build a #RAG app to chat with your favorite blogs blog.dagworks.io/p/from-blog-to… This post covers #opensource tooling like Burr Framework dltHub LanceDB OpenAI OpenTelemetry !

LanceDB seems to have implemented batched embeddings with oai when your query_type is "vector"? Firing off maybe 200 queries at one time and suddenly 183 return simultaneously is magical lol.

Learn how to build a multi-document agentic RAG system for automotive needs using LanceDB! This walkthrough shows you how to create an AI agent that can diagnose car issues, suggest fixes, and manage maintenance schedules. 🔧 Set up vector databases with LanceDB for efficient

lu.ma/roev9rc6 Welcome to the first event in the Compound AI Systems meetup series for Fall/Winter 24/25! Present to you by LanceDB and Databricks We bring together people from research and industry to highlight Composable Data Systems, AI/ML integrations, and Open

🙌Ready for next Tuesday’s geek out session with us? 📢 It’s time for us to review all the speaker lineups: changhiskhan - CEO Cofounder from LanceDB, Erik Bernhardsson - Founder CEO from Modal , and Kobie Crawford - Developer Advocate from Databricks for our 1st meetup in the

I ran some experiments with ColPali & LanceDB. Here are some takeaways: - Vision retrievers represent a significant step function change in the document retrieval process. blog.lancedb.com/late-interacti…

Revealing our 2nd talk at the upcoming meet up, by Kobie Crawford from Databricks Register now for a spot! lu.ma/roev9rc6

Amazing work on how to speed up search with the ColPali model on top LanceDB !