3.2.1.2. Sampling and Generation Parameters

3.2.1.2. Sampling and Generation Parameters

Sampling parameters control the randomness and creativity of generated responses. Understanding these is critical for the exam.

Testable Pattern:

from openai import AzureOpenAI

client = AzureOpenAI(azure_endpoint=endpoint, api_key=key, api_version="2024-08-01-preview")
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello"}],
    temperature=0.7
)
answer = response.choices[0].message.content

Error Handling Pattern:

from openai import AzureOpenAI, APIError, RateLimitError

try:
    response = client.chat.completions.create(model="gpt-4o", messages=messages)
except RateLimitError:
    # Back off and retry - quota exceeded
    time.sleep(60)
except APIError as e:
    # Log error details: e.status_code, e.message
    logging.error(f"API error: {e.status_code}")

CLI Equivalent (REST):

curl -X POST "https://{resource}.openai.azure.com/openai/deployments/{deployment}/chat/completions?api-version=2024-08-01-preview" \
  -H "Content-Type: application/json" \
  -H "api-key: {key}" \
  -d '{"messages": [{"role": "user", "content": "Hello"}]}'

⚠️ Exam Trap: temperature=0 gives deterministic output; temperature=1.0+ increases creativity.