Mirascope Integration
Mirascope (GitHub) is a Python toolkit for building LLM applications.
Developing LLM-applications with Mirascope feels just like writing the Python code you’re already used to. Python Toolkit for LLMs: Mirascope simplifies the development of applications using Large Language Models (LLMs) by providing an intuitive interface similar to standard Python coding practices.
The Mirascope team recorded a video with a short demo of how to use Mirascope with Langfuse.
How to use Mirascope with Langfuse
from mirascope.langfuse import with_langfuseMirascope automatically passes the Langfuse observe() decorator to all relevant functions within Mirascope via its with_langfuse decorator.
Example
Call
This is a basic call example but will work with all of Mirascope’s call functions, call, stream, call_async, stream_async.
import os
from mirascope.langfuse import with_langfuse
from mirascope.anthropic import AnthropicCall
os.environ["LANGFUSE_SECRET_KEY"] = "sk-lf-..."
os.environ["LANGFUSE_PUBLIC_KEY"] = "pk-lf-..."
os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
@with_langfuse
class BookRecommender(AnthropicCall):
prompt_template = "Please recommend some {genre} books"
genre: str
recommender = BookRecommender(genre="fantasy")
response = recommender.call() # this will automatically get logged with langfuse
print(response.content)
#> Here are some recommendations for great fantasy book series: ...This will give you:
- A trace around the AnthropicCall.call() that captures items like the prompt template, and input/output attributes and more.
- Human-readable display of the conversation with the agent
- Details of the response, including the number of tokens used
Extract
Mirascope’s extraction functionality is built on top of Pydantic and offers a convenient
extractmethod on extractor classes to extract structured information from LLM outputs. This method leverages tools (function calling) to reliably extract the required structured data. (Docs)
import os
from mirascope.langfuse import with_langfuse
from mirascope.openai import OpenAIExtractor
os.environ["LANGFUSE_SECRET_KEY"] = "sk-lf-..."
os.environ["LANGFUSE_PUBLIC_KEY"] = "pk-lf-..."
os.environ["LANGFUSE_HOST"] = "https://cloud.langfuse.com"
class TaskDetails(BaseModel):
description: str
due_date: str
priority: Literal["low", "normal", "high"]
@with_langfuse
class TaskExtractor(OpenAIExtractor[TaskDetails]):
extract_schema: Type[TaskDetails] = TaskDetails
prompt_template = """
Extract the task details from the following task:
{task}
"""
task: str
task = "Submit quarterly report by next Friday. Task is high priority."
task_details = TaskExtractor(
task=task
).extract() # this will be logged automatically with langfuse
assert isinstance(task_details, TaskDetails)
print(task_details)
# > description='Submit quarterly report' due_date='next Friday' priority='high'This will give you the same view as you would get from using the OpenAI integration. We will be adding more extraction support for other providers soon.
