[
https://issues.apache.org/jira/browse/CAMEL-21539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17913279#comment-17913279
]
James Netherton commented on CAMEL-21539:
-----------------------------------------
I have a POC where I can store and query an embedding created by Langchain4j.
It works in a similar way to the Langchain4j Infinispan embedding store. I
added options so you can configure the vector dimension, similarity and the
distance value to use in kNN searches.
Example usage with langchain4j-embeddings:
{code:java}
// Store an embedding
from("direct:put")
.to("langchain4j-embeddings:test")
.transform(new DataType("infinispan:embeddings"))
.to("infinispan:test%s?embeddingStoreDimension=3");
// Query embeddings using a pre-baked kNN query
from("direct:query")
.to("langchain4j-embeddings:test")
.setHeader(InfinispanConstants.OPERATION).constant(InfinispanOperation.QUERY)
.transform(new DataType("infinispan:embeddings"))
.to("infinispan:test?embeddingStoreDimension=3");
{code}
[~zbendhiba] WDYT?
> Vector Database capabilities - Infinispan
> -----------------------------------------
>
> Key: CAMEL-21539
> URL: https://issues.apache.org/jira/browse/CAMEL-21539
> Project: Camel
> Issue Type: New Feature
> Reporter: Zineb Bendhiba
> Assignee: James Netherton
> Priority: Major
>
> Implement Vector search capabilities for Infinispan.
> Also add the langchain4j Embeddings DataFormat for Infinispan + an
> Integration test in the langchain4j embeddings directory
> https://github.com/apache/camel/tree/main/components/camel-ai/camel-langchain4j-embeddings/src/test/java/org/apache/camel/component/langchain4j/embeddings
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
