Further development of the library

[ChristianWeyer]

Hi - great effort!

Are you planning to further develop the lib?
The original semantic-router has moved on since then.

Thanks!

[racoolstudio]

Hello @ChristianWeyer,

Thank you so much for your kind words and for bringing this to my attention! I truly appreciate your support and feedback.

Yes, I do have plans to further develop the library. I’m aware that the original semantic-router has evolved, and I’m exploring ways to align or enhance the library to provide more value. If you have any specific features or improvements in mind, I’d be happy to consider them as part of the roadmap.

Looking forward to hearing your thoughts!

Thank you once again !

[ChristianWeyer]

What I think is a strong and much-needed scenario is to use local embedding models for routing.

[racoolstudio]

Hello @ChristianWeyer,

Thank you for the insightful suggestion!

I’m definitely considering the idea of using a local embedder. Instead of relying on the SmartComponent Model, I could explore one of the models available on Hugging Face, such as sentence-transformers/all-MiniLM-L6-v2, or an even better alternative. By downloading the ONNX file and potentially implementing a custom tokenizer, it would optimize the process and ensure the solution remains as lightweight as possible.

Thank you once again for your valuable input!

[ChristianWeyer]

I think the model needs to be configurable. E.g., with the original semantic-router, I am using FastEmbed encoder with the intfloat/multilingual-e5-large model. Super capable and super fast.

[racoolstudio]

Yes, I perfectly understand you.
But C sharp is very limited even the use of Semantic Kernel is still limited.
What I can do is try to implement the local embedder that would use this same model as you stated above and see how it goes but I have to configure everything myself.