wenjin272 opened a new issue, #426:
URL: https://github.com/apache/flink-agents/issues/426

   ### Search before asking
   
   - [x] I searched in the 
[issues](https://github.com/apache/flink-agents/issues) and found nothing 
similar.
   
   ### Description
   
   Currently, Python and Java both have the implementation 
`VectorStoreLongTermMemory` for Long-Term Memory, which are logically identical.
   
   Maintaining two logically identical implementations in both Python and Java 
is difficult in the long run. We can take inspiration from short-term memory: 
maintain the implementation only on the Java side, and have the Python side 
uniformly invoke the Java implementation, performing parameter and result 
conversion as needed.
   
   Previously, we couldn't do this because, unlike Short-Term Memory—which 
relies solely on Java-native Flink state—Long-Term Memory depends on 
language-bound resources such as VectorStore, Embedding Model, and Chat Model. 
However, with support for cross-language invocation of resources now in place, 
this has become possible.
   
   Since this unification requires significant refactoring of the existing LTM 
implementation, and we expect this refactoring to be API-compatible, we will 
implement it in next version 0.3.
   
   ### Are you willing to submit a PR?
   
   - [x] I'm willing to submit a PR!


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to