hard20 pts · first answer countsragfine-tuningmodel-selection

Fresh facts plus a fixed style

A team needs answers in a very specific house style, using product facts that change every week. Which approach fits best?

A.Use retrieval for the changing facts, and handle style with prompting and examples, or fine-tuning if prompting is not enoughB.Fine-tune the model every week on the new facts onlyC.Put all product facts in a fixed system prompt foreverD.Rely on the model's training data for the facts

← API key in the system prompt Long agent run fills context →