Your question is Constructing a Fine-Tuning Dataset. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How would you construct a high-quality fine-tuning dataset for Meituan ML models?