2 ms·
I'm surprised you landed on using o3 as the judge - we found it way too expensive. I use llm as a judge for generating color variations of products, definitely
by shmoogy 1y ago
I'm surprised you landed on using o3 as the judge - we found it way too expensive. I use llm as a judge for generating color variations of products, definitely hoping for some improvements - it can be brutal to get non hallucinated features along with proper final rendering.
- omneity 1y agoHave you tried open weights vision models such as Qwen VL, MiniCPM, PaliGemma...? I'm also curious how usable are simpler vision models such as Florence in case you explored this direction.
- palashshah 1y agowe're currently in the process of doing this. i think something that could potentially work is to iterate upon the initial image composition / structure using cheaper models, and then upscale at the end. this way you're saving on that iteration cost, but eventually land on a higher-scale image.
- shmoogy 1y agoI actually haven't but nova from Amazon was surprisingly good at things like bounding boxes compared to some others You kind of have to test and measure so many different aspects to get the best at specific tasks Thanks for the idea