Data/ML/AI
Fine-Tuning
Further training an existing pre-trained model on a smaller, specific dataset so it performs better on a particular task or domain.
Think of it like
Like sending a generally-skilled new hire through a specialized in-house training course, rather than hiring a brand-new specialist from scratch.
Junior or senior?
Junior sounds like
Reaches for fine-tuning as the default approach.
Senior sounds like
Can explain what made fine-tuning the right call instead of cheaper, faster prompting or RAG.
Ask them
“What made you fine-tune a model instead of solving it with prompting or retrieval?”