Microsoft Exam AI-102 Topic 3 Question 60 Discussion

Actual exam question for Microsoft's AI-102 exam

Question #: 60
Topic #: 3

[All AI-102 Questions]

You build a chatbot that uses the Azure OpenAI GPT 3.5 model.

You need to improve the quality of the responses from the chatbot. The solution must minimize development effort.

What are two ways to achieve the goal? Each correct answer presents a complete solution.

NOTE: Each correct answer is worth one point.

AFine-tune the model.

BProvide grounding content.

CAdd sample request/response pairs.

DRetrain the language model by using your own data.

ETrain a custom large language model (LLM).

Show Suggested Answer

Suggested Answer: B, C

by Merlyn at Apr 09, 2024, 12:28 PM

Limited Time Offer

25%

Off

Get Premium AI-102 Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

Submit Cancel

Alishia

6 days ago

I was eyeing that option to train a custom large language model (LLM), but that just seems way too ambitious for this particular use case. Talk about going overboard!

upvoted 0 times

...

Jamal

6 days ago

I'm leaning towards the grounding content and sample request/response pairs options. They seem like the most straightforward ways to improve the chatbot's responses without too much hassle.

upvoted 0 times

...

Latosha

7 days ago

I was considering retraining the language model using our own data, but that sounds like it could be a ton of work. Plus, I'm not sure if that's really minimizing the development effort.

upvoted 0 times

...

Agreed. I think fine-tuning the model or adding sample request/response pairs are our best bets here. Though I do wonder if we can sneak in a few dad jokes while we're at it, you know, to really make the chatbot shine.

upvoted 0 times

...

Buffy

8 days ago

Wait, can we just train a custom LLM from scratch? That sounds like it would give us the most control over the model's behavior. Although, it might take a lot of time and resources.

upvoted 0 times

...

William

9 days ago

I'm a little worried that fine-tuning the model might be too much work. Wouldn't it be easier to just add some sample data and let the model learn from that?

upvoted 0 times

...

Bettye

9 days ago

Haha, yeah, I don't think we want to become language model experts overnight. Let's stick to the simpler solutions, yeah?

upvoted 0 times

...

Louis

10 days ago

Adding sample request/response pairs seems like a smart move. We can feed the model some examples of the kind of interactions we want, and it should be able to learn from that.

upvoted 0 times

...

Denny

12 days ago

I think providing grounding content could be a good option. We can give the model some context and background information to help it generate more relevant and coherent responses.

upvoted 0 times

...

Katy

14 days ago

Wow, this is a tricky one! I'm not sure if I can fine-tune the model without a ton of data and processing power. And retraining the whole language model seems like overkill for just improving the chatbot's responses.

upvoted 0 times

...

Microsoft Exam AI-102 Topic 3 Question 60 Discussion

Contribute your Thoughts:

Alishia

Jamal

Latosha

Daron

Buffy

William

Bettye

Louis

Denny

Katy