13. LLM Cheatsheet

OpenAI API

python
client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Hi"}]
)

python
pipe = pipeline("text-generation", model="gpt2")
pipe("The sky is")

python
prompt = PromptTemplate.from_template("Tell me a {adj} joke")
chain = prompt | llm