Generative AI Leader Fundamentals of Generative AI Practice Question
A data scientist fine-tunes a foundation model on customer support transcripts. After evaluation, the model's responses are too formal. Which adjustment during fine-tuning is most likely to make responses more conversational?
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Include examples of informal customer interactions in the fine-tuning data.
The training data directly influences the tone and style of model outputs. Including examples of informal conversations in the fine-tuning dataset teaches the model the desired conversational tone. Other options affect training dynamics but not the style.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase the batch size to stabilize training.
Why it's wrong here
Batch size governs gradient stability and throughput, not the stylistic register the model learns; it cannot shift tone from formal to conversational. It is tempting because larger batches do stabilise noisy training runs, and that would be the right lever when loss curves oscillate rather than when output style needs changing.
- ✗
Decrease the number of fine-tuning steps to prevent overfitting.
Why it's wrong here
Reducing steps limits how strongly the model absorbs the transcripts, but formality stems from the training examples' register, not from overfitting; fewer steps simply underfits. It tempts because early stopping is the standard remedy for overfitting, which would be correct if validation loss were rising.
- ✓
Include examples of informal customer interactions in the fine-tuning data.
Why this is correct
Fine-tuning adapts a model's style to its training distribution, so the formal tone reflects the transcripts used. Adding informal customer interaction examples shifts that distribution toward conversational phrasing, directly satisfying the requirement to make responses less formal without changing architecture or decoding parameters.
- ✗
Use a higher learning rate for faster adaptation.
Why it's wrong here
Learning rate controls how far weights move per update, not which stylistic patterns the data teaches; a higher rate risks divergence without altering register. It tempts because faster adaptation suits domain shift, and that would be correct if the model needed to absorb new vocabulary rather than change tone.
Go deeper
Related to this question
About these practice questions
One of 1,008 original Generative AI Leader practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.