Discuss data (and model) update cycle for operational deployment

soilwise-he / natural-language-querying

Application component that provides Natural Language Querying (NLQ) services, making knowledge stored in a graph database accessible for e.g. a ChatBot UI.

MIT License

0 stars 0 forks source link

Discuss data (and model) update cycle for operational deployment #10

Open robknapen opened 5 months ago

robknapen commented 5 months ago

New data will be harvested periodically which will somehow be reflected in the NLQ component. E.g. it might need to updating of embeddings in a vector store.

New versions of LLMs are made available frequently at the moment, updating of what is used by the NLQ component might also be needed accordingly. I.e. for security or model toxicity reasons.

BerkvensNick commented 4 months ago

Good point, I think this can only be discussed when we have defined a more clear use case we want to develop/set up