weaviate
weaviate copied to clipboard
weaviate
Weaviate is an open-source vector database that stores both objects and vectors, allowing for the combination of vector search with structured filtering with the fault tolerance and scalability of a c...
**Tl;dr** Vectorize a class with the average vectors of a cross-referenced class. **Motivation** Averaging dense vectors from Deep Learning models preserves the semantics of the cluster. We see applications for...
Hi, GPT-3 embeddings can be called upon a fine tuned model. But I'm not sure the fine tuned model can be set on the class definition, as described in the...
## Background The current condensor loads an uncondensed commit log into memory and then writes the condensed version back to disk. This creates two problems: 1. Either we have massive...
Fix memory leaks Improve memory allocations using new strategy Improve implementation of the list pool and RC
Hello, I want to start up weaviate on the localhost for basic testing and need some help in determining what configurations are needed to do this. I want to start...
Based on [this](https://github.com/semi-technologies/weaviate/issues/2078#issuecomment-1205797943) comment.
## Reproduce + Actual Behavior https://link.semi.technology/3vDpdpK ``` vector search: search index article: shard article_SZAfi4UN1F77: vector search: cannot vector-search on a class not vector-indexed ``` ## Desired Behavior * If all...
This is to support the "_online reloading of a class_" pattern In production, in large-scale use of Weaviate, one would need to implement an online data reloading strategy for large...
It would be great to be able to ingest data into a Weaviate with a Spark DataFrameWriter. This would allow easier integration for those who are currently using Lakehouse architectures...
