LoslegenKostenlos starten

Query a Foundation Model

Let's use what we learned about the difference between "SYSTEM" and "USER" Chat Message Roles to send different prompts to the AI agent based on the Chat Message role.

Diese Übung ist Teil des Kurses

<Kurs>Databricks with the Python SDK</Kurs>
Kurs ansehen

Übungsanweisungen

  • Tell the AI Agent to act like a high school computer science teacher when answering queries using the prompt ai_agent_context.
  • Ask the AI agent when to use For loops using the user_prompt.

Interaktive praktische Übung

Versuche dich an dieser Übung, indem du diesen Beispielcode vervollständigst.

from databricks.sdk import WorkspaceClient
from databricks.sdk.service.serving import ChatMessage, ChatMessageRole
w = WorkspaceClient()
user_prompt = "What is a For loop and When should I use it?"
ai_agent_context = "You are a high school computer science teacher"
response = w.serving_endpoints.query(
    name="databricks-meta-llama-3-3-70b-instruct",
    messages=[
        ChatMessage( 
          	# Tell the AI Agent to act like a high school computer science teacher when answering queries
            role=ChatMessageRole.SYSTEM, content=____
        ),
        ChatMessage( 
          	# Ask the AI agent when to use For loops
            role=ChatMessageRole.USER, content=____
        ),
    ],
    max_tokens=250)
print(f"RESPONSE:\n{response.choices[0].message.content}")
Code bearbeiten und ausführen