What does “Retrieve engine” mean? I understand that when I send completion request (using Python), on server side an engine is instantiated for that task… I that correct?
If so, then how long my engine instance is active? Can I work with the same instance and make new requests to it?
I’m asking this because I see that simple chat completions are extremely dependent on the Q&A chain that preceded the result plus some randomness. The completion process is semi-stochastic so the current instance is “one of a kind”. I would like to work with an instance that I have created and “conditioned” for very specific tasks - the instance which works the best.
Sorry if I’m rambling - I got access three days ago, and haven’t slept very much since…
1 Like
That’s correct. There’s no difference in calling an engine with a specific prompt today vs. tomorrow.
1 Like
Ok, I see. Thanks for the clarification!
1 Like