Logfire

Logfire, a new tool from Pydantic, is built on OpenTelemetry. As Mirascope is also built on Pydantic, it's appropriate for us to ensure seamless integration with them.

How to use Logfire with Mirascope

from mirascope.logfire import with_logfire

with_logfire is a decorator that can be used on all Mirascope classes to automatically log both Mirascope calls and also all our supported LLM providers.

Examples

Call

This is a basic call example but will work with all our call functions, call, stream, call_async, stream_async.

import logfire
from mirascope.logfire import with_logfire
from mirascope.anthropic import AnthropicCall

logfire.configure()


@with_logfire
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 logfire
print(response.content)
#> Here are some recommendations for great fantasy book series: ...

This will give you:

• A span around the AnthropicCall.call() that captures items like the prompt template, templating properties and fields, and input/output attributes
• Human-readable display of the conversation with the agent
• Details of the response, including the number of tokens used

Extract

Since Mirascope is built on top of Pydantic, you can use the Pydantic plugin to track additional logs and metrics about model validation, which you can enable using the pydantic_plugin configuration.

This can be particularly useful when extracting structured information using LLMs:

from typing import Literal, Type

import logfire
from mirascope.logfire import with_logfire
from mirascope.openai import OpenAIExtractor
from pydantic import BaseModel

logfire.configure(pydantic_plugin=logfire.PydanticPlugin(record="all"))


class TaskDetails(BaseModel):
    description: str
    due_date: str
    priority: Literal["low", "normal", "high"]


@with_logfire
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 logfire
assert isinstance(task_details, TaskDetails)
print(task_details)
#> description='Submit quarterly report' due_date='next Friday' priority='high'

This will give you:

• Tracking for validation of Pydantic models
• A span around the OpenAIExtractor.extract() that captures items like the prompt template, templating properties and fields, and input/output attributes
• Human-readable display of the conversation with the agent including the function call
• Details of the response, including the number of tokens used

FastAPI

You can take advantage of existing instruments from logfire and integrate it with Mirascope.

import os
from typing import Type

import logfire
from fastapi import FastAPI
from pydantic import BaseModel

from mirascope.logfire import with_logfire
from mirascope.openai import OpenAIExtractor

os.environ["OPENAI_API_KEY"] = "YOUR_OPENAI_API_KEY"

app = FastAPI()
logfire.configure()
logfire.instrument_fastapi(app)


class Book(BaseModel):
    title: str
    author: str


@with_logfire
class BookRecommender(OpenAIExtractor[Book]):
    extract_schema: Type[Book] = Book
    prompt_template = "Please recommend a {genre} book."

    genre: str


@app.post("/")
def root(book_recommender: BookRecommender) -> Book:
    """Generates a book based on provided `genre`."""
    return book_recommender.extract()

This will generate a well-structured hierarchy. This way, you can view your API calls, Mirascope models, and LLM calls all in one place with just a few lines of code.

RAG

from mirascope.chroma import ChromaSettings, ChromaVectorStore
from mirascope.logfire import with_logfire
from mirascope.openai import OpenAIEmbedder
from mirascope.rag import TextChunker

@with_logfire
class MyStore(ChromaVectorStore):
    embedder = OpenAIEmbedder()
    chunker = TextChunker(chunk_size=1000, chunk_overlap=200)
    index_name = "index-0001"
    client_settings = ChromaSettings()

store = MyStore()
store.add("some data") # this will automatically get logged with logfire

Integrations with other providers

At present, Logfire includes instrument_openai and instrument_fastapi. As the Logfire team introduces more instruments, the Mirascope team will also update with_logfire to incorporate these instruments. Meanwhile, Mirascope ensures Logfire integration with all of its providers. Here is an example using Anthropic:

import logfire

from mirascope.anthropic import AnthropicCall
from mirascope.logfire import with_logfire

logfire.configure()

@with_logfire
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 logfire
print(response.content)