The creative story

Your prompt determines what the AI says. Parameters determine how it says it. Temperature controls randomness, top_p controls vocabulary diversity, and max_tokens caps the response length. Getting these right is the difference between a creative story and a repetitive one.

LLM parameter effects

How temperature and max_tokens affect the output.

service.py

python

# These parameters are passed to the LLM provider
response = await llm_provider.generate(
    prompt,
    temperature=0.8,  # 0.0 = deterministic, 1.0+ = creative
    max_tokens=800    # Limits story length
)

Temperature 0.8 is a good default for creative tasks. Lower it for factual queries (0.2) or raise it for brainstorming (1.0+). max_tokens prevents runaway responses. The same parameters work with generate_stream, the streaming call you learn next.

Matching exercise: Match parameters to their effects

Loading practice…