AI-102 Question 282
Select 3You are developing a client application that will consume a pre-trained Azure OpenAI GPT-4 language model. The application requires sending user input to the model and receiving responses in real-time. Which steps must you implement to successfully interact with the language model?
- A
Obtain the endpoint URL and access key from the Azure OpenAI resource.
- B
Use the Azure CLI to directly send input to the language model without additional setup.
- C
Send a POST request to the model's endpoint with the API key included in the request headers.
- D
Format the request payload to include user input and any required parameters, such as temperature or max tokens.
- E
Enable Azure Kubernetes Service (AKS) for deploying the language model before consuming it.
Show answer and explanation
Correct answers: A, C, D
Explanation
To consume a language model from a client application, you need to authenticate using the endpoint URL and access key, send a properly formatted POST request, and include user input along with configurable parameters. Azure CLI and AKS are not part of the required steps for interacting with the Azure OpenAI service in this scenario.
- A. Correct.
Correct. To interact with the Azure OpenAI language model, you need the endpoint URL and access key, which are provided when you set up the Azure OpenAI resource.
- B. Incorrect.
Incorrect. While the Azure CLI is useful for managing Azure resources, it is not used to send real-time input to the language model.
- C. Correct.
Correct. A POST request with the API key in the headers is required to authenticate and interact with the language model's endpoint.
- D. Correct.
Correct. The request payload must include user input and other important parameters like temperature or max tokens to customize the model's behavior.
- E. Incorrect.
Incorrect. There is no requirement to deploy the language model using Azure Kubernetes Service (AKS) for client applications to consume it. The model is already hosted as part of the Azure OpenAI service.