Skip to content

armgit / dspy

Public Imported from GitHub
0

DSPy: The framework for programming—not prompting—language models

Python

Add test for custom rationale type check

Herumb Shandilyabcaf13d834c0
diff --git a/tests/functional/test_functional.py b/tests/functional/test_functional.py
index db39e069f..b8c341b3c 100644
--- a/tests/functional/test_functional.py
+++ b/tests/functional/test_functional.py
@@ -619,6 +619,48 @@ def test_list_input2():
Proposed Signature: Output""")
+def test_custom_rationale_type():
+ class Question(pydantic.BaseModel):
+ value: str
+
+ class QuestionSignature(dspy.Signature):
+ topic: str = dspy.InputField()
+ question: Question = dspy.OutputField()
+
+ rationale_type = dspy.OutputField(
+ prefix="Custom Reasoning: Let's break this down. To generate a question about",
+ desc="${topic}, we should ...",
+ )
+
+ program = TypedChainOfThought(QuestionSignature, rationale_type=rationale_type)
+
+ expected = "What is the speed of light?"
+ lm = DummyLM(["Thoughts", f'{{"value": "{expected}"}}'])
+ dspy.settings.configure(lm=lm)
+
+ output = program(topic="Physics")
+
+ assert isinstance(output.question, Question)
+ assert output.question.value == expected
+
+ assert lm.get_convo(-1) == textwrap.dedent("""\
+ Given the fields `topic`, produce the fields `question`.
+
+ ---
+
+ Follow the following format.
+
+ Topic: ${topic}
+ Custom Reasoning: Let's break this down. To generate a question about ${topic}, we should ...
+ Question: ${question}. Respond with a single JSON object. JSON Schema: {"properties": {"value": {"title": "Value", "type": "string"}}, "required": ["value"], "title": "Question", "type": "object"}
+
+ ---
+
+ Topic: Physics
+ Custom Reasoning: Let's break this down. To generate a question about Thoughts
+ Question: {"value": "What is the speed of light?"}""")
+
+
def test_generic_signature():
T = TypeVar("T")