dspy-prompting（dspy 提示）

DSPy Declarative Language Model Programming

Build AI systems with automatic prompt optimization from Stanford NLP.

When to Use

• Building complex AI systems with multiple components
• Programming LMs declaratively instead of manual prompting
• Optimizing prompts automatically using data-driven methods
• Creating modular AI pipelines that are maintainable
• Building RAG systems, agents, or classifiers with better reliability

Quick Start

pip install dspy

Basic Question Answering

import dspy

lm = dspy.Claude(model="claude-sonnet-4-5-20250929")
dspy.settings.configure(lm=lm)

# Define a signature (input -> output)
class QA(dspy.Signature):
    """Answer questions with short factual answers."""
    question = dspy.InputField()
    answer = dspy.OutputField(desc="often between 1 and 5 words")

qa = dspy.Predict(QA)
response = qa(question="What is the capital of France?")
print(response.answer)  # "Paris"

Chain of Thought Reasoning

class MathProblem(dspy.Signature):
    """Solve math word problems."""
    problem = dspy.InputField()
    answer = dspy.OutputField(desc="numerical answer")

cot = dspy.ChainOfThought(MathProblem)
response = cot(problem="If John has 5 apples and gives 2 to Mary, how many does he have?")
print(response.rationale)  # Shows reasoning steps
print(response.answer)     # "3"

Core Modules

ReAct Agent

from dspy.predict import ReAct

class SearchQA(dspy.Signature):
    """Answer questions using search."""
    question = dspy.InputField()
    answer = dspy.OutputField()

def search_tool(query: str) -> str:
    """Search Wikipedia."""
    return results

react = ReAct(SearchQA, tools=[search_tool])
result = react(question="When was Python created?")

Automatic Optimization

BootstrapFewShot

from dspy.teleprompt import BootstrapFewShot

trainset = [
    dspy.Example(question="What is 2+2?", answer="4").with_inputs("question"),
    dspy.Example(question="What is 3+5?", answer="8").with_inputs("question"),
]

def validate_answer(example, pred, trace=None):
    return example.answer == pred.answer

optimizer = BootstrapFewShot(metric=validate_answer, max_bootstrapped_demos=3)
optimized_qa = optimizer.compile(qa, trainset=trainset)

MIPRO Optimizer

from dspy.teleprompt import MIPRO

optimizer = MIPRO(
    metric=validate_answer,
    num_candidates=10,
    init_temperature=1.0
)

optimized_cot = optimizer.compile(cot, trainset=trainset, num_trials=100)

Multi-Stage Pipeline

class MultiHopQA(dspy.Module):
    def __init__(self):
        super().__init__()
        self.retrieve = dspy.Retrieve(k=3)
        self.generate_query = dspy.ChainOfThought("question -> search_query")
        self.generate_answer = dspy.ChainOfThought("context, question -> answer")

    def forward(self, question):
        search_query = self.generate_query(question=question).search_query
        passages = self.retrieve(search_query).passages
        context = "\n".join(passages)
        answer = self.generate_answer(context=context, question=question).answer
        return dspy.Prediction(answer=answer, context=context)

Structured Output

from pydantic import BaseModel, Field

class PersonInfo(BaseModel):
    name: str = Field(description="Full name")
    age: int = Field(description="Age in years")
    occupation: str = Field(description="Current job")

class ExtractPerson(dspy.Signature):
    """Extract person information from text."""
    text = dspy.InputField()
    person: PersonInfo = dspy.OutputField()

extractor = dspy.TypedPredictor(ExtractPerson)
result = extractor(text="John Doe is a 35-year-old software engineer.")

LLM Providers

# Anthropic
lm = dspy.Claude(model="claude-sonnet-4-5-20250929")

# OpenAI
lm = dspy.OpenAI(model="gpt-4")

# Local (Ollama)
lm = dspy.OllamaLocal(model="llama3.1", base_url="http://localhost:11434")

dspy.settings.configure(lm=lm)

Save and Load

# Save optimized module
optimized_qa.save("models/qa_v1.json")

# Load later
loaded_qa = dspy.ChainOfThought("question -> answer")
loaded_qa.load("models/qa_v1.json")

vs Alternatives

Choose DSPy when:

• You have training data or can generate it
• Need systematic prompt improvement
• Building complex multi-stage systems

Resources

• Docs: https://dspy.ai
• GitHub: https://github.com/stanfordnlp/dspy
• Paper: "DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines"