AI-102 exam dumps

AI-102 practice question 465 of 493

Designing and Implementing a Microsoft Azure AI Solution. Professional level, Microsoft. Free question with the correct answer and a full explanation.

AI-102 Question 465

Select 4

Your company wants to create a customer support chatbot that can handle both text and image-based inputs, such as providing product information from text queries or identifying products from uploaded images. You decide to use Azure OpenAI and its large multimodal models for this purpose. Which steps should you take to implement this solution effectively?

  1. A

    Deploy the GPT-4 model with vision capabilities in Azure OpenAI.

  2. B

    Enable multimodal input support during the deployment configuration.

  3. C

    Use the Azure OpenAI Service REST API to send both text and image inputs to the model.

  4. D

    Fine-tune the GPT-4 model to support multimodal inputs.

  5. E

    Integrate the Azure OpenAI Service with other Azure services like Cognitive Search to enhance functionality.

Show answer and explanation

Correct answers: A, B, C, E

Explanation

To implement a multimodal solution with Azure OpenAI, you need to deploy a large multimodal model like GPT-4 with vision capabilities and configure it to support multimodal inputs. Sending inputs (text and images) via the REST API is essential for interaction. Fine-tuning is unnecessary as the model already supports multimodal inputs. Integrating with additional Azure services, such as Cognitive Search, can further enhance the solution by providing augmented capabilities like retrieving relevant product information.

  • A. Correct.

    Correct. GPT-4 with vision capabilities is a large multimodal model in Azure OpenAI and supports both text and image inputs.

  • B. Correct.

    Correct. You need to enable multimodal input support during deployment to ensure the model can process both text and image inputs.

  • C. Correct.

    Correct. The Azure OpenAI Service REST API allows you to send multimodal inputs (text and images) to the model for processing.

  • D. Incorrect.

    Incorrect. Fine-tuning is not required for GPT-4 with vision capabilities as it already supports multimodal inputs out-of-the-box.

  • E. Correct.

    Correct. Integrating with services like Azure Cognitive Search can enhance the chatbot's ability to retrieve and process additional information, improving its overall functionality.

Timed practice exam

Take a AI-102 practice test under exam conditions

40 questions in 120 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam