Just a quick stupid question for confirmation. Are we only allowed to use Gemini Models or can we use Gemini models along with other gen ai models?
Coz I have already recorded a video (to balance out costings and quality I have used several models - including gemini-1.0-pro [a greater portion uses this] and gemini-1.5-flash-latest [used this for parts where more reasoning is required]). Since, my app also uses other services and I have exhausted my GPU quota, so, would be a little difficult to re-run the entire thing with all Gemini.
We discussed this a bit here: Another API in the competition
While Gemini has to be the core focus of your work, you’re certainly allowed to use other models where they make sense.
Your reasoning is a bit odd, however. There is no specific “GPU quota” for Gemini, although there is a limit on the number of tokens per minute or day. Those should reset, however.
Use the best model for the project. Remember that you’re submitting code as well for them to review.
1 Like
Thank you for the response!
Oh my! I should have been a bit clear, when I say GPU quota, I don’t mean Gemini, I mean some other open-source models that I have hosted for other tasks on GPUs.
Yes! I’ll submit a zip. But since its a web application and which Gemini model to use is decided during runtime based on config saved in db, one might not be able to account for all the Gemini Models that we are using (as mentioned in my main post). But yes, it would be clearly visible that we are using Gemini API, if that all it asks.
Just one more question: When submitting the code do we need to submit all services. I have 2 frontends and 2 backends.
I would assume yes as you are partially being rated on your code, so leaving out most of it would not be good for your “score”.
1 Like