[
https://issues.apache.org/jira/browse/SOLR-17927?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18027934#comment-18027934
]
Ishan Chattopadhyaya commented on SOLR-17927:
---------------------------------------------
bq. + 1 for renaming neural search to vector search.
+1
bq. In regard to the 'llm' module, I'm a bit more dubious:
bq. The intent of that module is to use langChain4j to expose LLM integrations
(starting from text vectorisation).
I don't mind naming the module langchain or rag or something like that. Even
language-models is better than LLM.
But, LLM is vague and unclear:
* There is no large language model in Solr
* There's model inference support, but that works for large and small language
models alike. For example sentence transformers like BERT are not referred to
as LLMs.
> Fix naming in vector search support
> -----------------------------------
>
> Key: SOLR-17927
> URL: https://issues.apache.org/jira/browse/SOLR-17927
> Project: Solr
> Issue Type: Improvement
> Components: vector-search
> Reporter: Ishan Chattopadhyaya
> Priority: Blocker
> Fix For: 10.0
>
>
> I propose we:
> * Rename llm (module) to "inference" or "embedding" module
> * Rename "neural search" to "vector search"
> Solr 10 is a good opportunity to do this.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
