Implement Custom Language Models AI-102 Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   In the Azure AI Language question answering portal, how can you test a knowledge base?

   • AUse the built-in test pane to submit questions and review answers
   • BTrain the knowledge base again
   • CExport the knowledge base to a JSON file
   • DDelete the knowledge base and recreate it
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A company has a custom language model that is critical to their customer service operations. They want to ensure they can restore the model quickly if the production deployment fails. What should they do?

   • AExport the trained model and store the backup in a secure location.
   • BRely on the cloud provider's automatic backups.
   • CKeep the training data only, since the model can be retrained from scratch.
   • DDo nothing, since the model is deployed and unlikely to fail.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A custom language model for a banking chatbot has high accuracy but is slow to respond, causing user frustration. The team wants to optimize the model's performance without sacrificing accuracy. What should they do first?

   • AAdd more entities to capture more details from user input.
   • BReview the model's training data for redundant or conflicting utterances and remove them, then retrain.
   • CIncrease the number of intents to make the model more specific.
   • DSwitch the model to a larger compute tier to improve speed.
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   A developer is integrating a custom language model into a web application. The app needs to send user input and receive the predicted intent and entities. What should the developer use to make the API call?

   • AThe training endpoint URL and the training dataset.
   • BThe model's export file and a local runtime.
   • CThe prediction endpoint URL and an authentication key.
   • DThe Azure portal's test console only.
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   In Azure AI Language, what is the first step to create a custom question answering solution?

   • APublish the knowledge base
   • BDeploy a model to an endpoint
   • CCreate a new project and connect it to a knowledge base
   • DTrain a language model
   Reveal Solution Discussion Select an answer first