AI 900 – Machine Learning Services \ Generative AI

[Brandon120]

Hello! I was hoping for some clarification on Azure Machine Learning.

First question I encountered was…
Your team is exploring generative AI workloads on Azure for the next project development. Your team is interested in using Azure’s capabilities to create and deploy Large Language Models (LLMs) for various applications, such as automated content generation, customer service chatbots, and natural language understanding.
Which Azure service can help the team deploy and manage large language models (LLMs)?

And this was part of the answer response…
“The option that says: Azure Machine Learning is incorrect. This service helps you in building, training, and deploying your custom machine learning models in Azure but doesn’t have any focus on generative AI workloads, unlike the Azure OpenAI Service.”

—

Then I had this question…
You work for an art gallery that aims to produce one-of-a-kind digital art pieces by blending famous paintings with modern styles.
The gallery intends to utilize generative AI models to accomplish this. As the AI specialist, you need to recommend the most suitable Azure service for this task.
Which Azure service would you suggest?

And this was part of the answer response…
“Hence, the correct answer is: Azure Machine Learning Service. This service is the most suitable for this task. It provides a comprehensive environment for training, deploying, and managing machine learning models. It supports deep learning frameworks like TensorFlow, PyTorch, and others, which are essential for developing and deploying generative AI models for producing digital art.”

—

I am confused as to whether Azure Machine Learning supports generative AI workloads or not given these two responses. Is there a difference between Azure Machine Learning and Azure Machine Learning Service?

[Brandon120]

Friendly bump on this question 😊

[JR-TutorialsDojo]

Hello Brandon120,

Thanks for the feedback.

There’s no difference between Azure Machine Learning and Azure Machine Learning Service. They refer to the same platform. The key distinction lies in what kind of generative AI workload you’re targeting:

• For deploying pretrained Large Language Models (LLMs) like GPT-4 for chatbots, content generation, or natural language understanding, the best fit is Azure OpenAI Service. It gives you direct access to powerful LLMs without needing to train them from scratch.
• For building custom generative models, especially in image generation, style transfer, or creative AI, Azure Machine Learning is ideal. It supports deep learning frameworks (like PyTorch or TensorFlow) and lets you train and deploy your own models.

Both responses are correct. They just apply to different types of generative AI:

• The first question is about LLMs, so Azure OpenAI Service is the most suitable answer.
• The second question is about generative art, which often involves custom models, so Azure Machine Learning is the better fit.

I hope this helps! Let us know if you need further assistance.

Best regards,
JR @ Tutorials Dojo

[Brandon120]

That makes a lot of sense now, thank you so much!!! 😊

[JR-TutorialsDojo]

You’re very welcome, Brandon. Always happy to assist.