AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Make a No-Lm Application That Works
If you were trying to make a no-lm application today, how much energy would you spend on fine-tuning the model itself versus getting the right prompt to send into the model? Can you give me some possible answers to that question? Give me an application that you've seen and how you would answer that for first common application. Yeah, so I would say in the context of let's look at no-mix auto-labeling engine, right? The first pass of this thing is prompt engineering that gets you quite far.