lmnr 0.4.7

Python SDK for Laminar AI

Laminar Python

OpenTelemetry log sender for Laminar for Python code.

Quickstart

python3 -m venv .myenv
source .myenv/bin/activate  # or use your favorite env management tool

pip install lmnr

And the in your main Python file

from lmnr import Laminar as L

L.initialize(project_api_key="<LMNR_PROJECT_API_KEY>")

This will automatically instrument most of the LLM, Vector DB, and related calls with OpenTelemetry-compatible instrumentation.

We rely on the amazing OpenLLMetry, open-source package by TraceLoop, to achieve that.

Project API key

Get the key from the settings page of your Laminar project (Learn more). You can either pass it to .initialize() or set it to .env at the root of your package with the key LMNR_PROJECT_API_KEY.

Instrumentation

In addition to automatic instrumentation, we provide a simple @observe() decorator, if you want more fine-grained tracing or to trace other functions.

Example

import os
from openai import OpenAI


from lmnr import observe, Laminar as L
L.initialize(project_api_key="<LMNR_PROJECT_API_KEY>")

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

@observe()  # annotate all functions you want to trace
def poem_writer(topic="turbulence"):
    prompt = f"write a poem about {topic}"
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": prompt},
        ],
    )
    poem = response.choices[0].message.content
    return poem

print(poem_writer(topic="laminar flow"))

Manual instrumentation

Our manual instrumentation is a very thin wrapper around OpenTelemetry's trace.start_span. Our wrapper sets the span into the active context. You don't have to explicitly pass the spans around, it is enough to just call L.start_span, and OpenTelemetry will handle the context management

from lmnr import observe, Laminar as L
L.initialize(project_api_key="<LMNR_PROJECT_API_KEY>")

def poem_writer(topic="turbulence"):
    
    span = L.start_span("poem_writer", topic) # start a span

    prompt = f"write a poem about {topic}"

    # OpenAI calls are still automatically instrumented with OpenLLMetry
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": prompt},
        ],
    )
    poem = response.choices[0].message.content
    # while within the span, you can attach laminar events to it
    L.event("event_name", "event_value")

    L.set_span_output(span, poem) # set an output
    
    # IMPORTANT: don't forget to end all the spans (usually in `finally` blocks)
    # Otherwise, the trace may not be sent/displayed correctly
    span.end()

    return poem

Sending events

You can send events in two ways:

• .event(name, value) – for a pre-defined event with one of possible values.
• .evaluate_event(name, evaluator, data) – for an event that is evaluated by evaluator pipeline based on the data.

Note that to run an evaluate event, you need to crate an evaluator pipeline and create a target version for it.

Read our docs to learn more about event types and how they are created and evaluated.

Example

from lmnr import Laminar as L
# ...
poem = response.choices[0].message.content

# this will register True or False value with Laminar
L.event("topic alignment", topic in poem)

# this will run the pipeline `check_wordy` with `poem` set as the value
# of `text_input` node, and write the result as an event with name
# "excessive_wordiness"
L.evaluate_event("excessive_wordiness", "check_wordy", {"text_input": poem})

Laminar pipelines as prompt chain managers

You can create Laminar pipelines in the UI and manage chains of LLM calls there.

After you are ready to use your pipeline in your code, deploy it in Laminar by selecting the target version for the pipeline.

Once your pipeline target is set, you can call it from Python in just a few lines.

Example use:

from lmnr import Laminar as L

L.initialize('<YOUR_PROJECT_API_KEY>')

result = l.run(
    pipeline = 'my_pipeline_name',
    inputs = {'input_node_name': 'some_value'},
    # all environment variables
    env = {'OPENAI_API_KEY': 'sk-some-key'},
)

Resulting in:

>>> result
PipelineRunResponse(
    outputs={'output': {'value': [ChatMessage(role='user', content='hello')]}},
    # useful to locate your trace
    run_id='53b012d5-5759-48a6-a9c5-0011610e3669'
)

Running offline evaluations on your data

You can evaluate your code with your own data and send it to Laminar using the Evaluation class.

Evaluation takes in the following parameters:

• name – the name of your evaluation. If no such evaluation exists in the project, it will be created. Otherwise, data will be pushed to the existing evaluation
• data – an array of EvaluationDatapoint objects, where each EvaluationDatapoint has two keys: target and data, each containing a key-value object. Alternatively, you can pass in dictionaries, and we will instantiate EvaluationDatapoints with pydantic if possible
• executor – the logic you want to evaluate. This function must take data as the first argument, and produce any output. *
• evaluators – evaluaton logic. List of functions that take output of executor as the first argument, target as the second argument and produce a numeric scores. Each function can produce either a single number or dict[str, int|float] of scores.

* If you already have the outputs of executors you want to evaluate, you can specify the executor as an identity function, that takes in data and returns only needed value(s) from it.

Example

from openai import AsyncOpenAI
import asyncio
import os

openai_client = AsyncOpenAI(api_key=os.environ["OPENAI_API_KEY"])

async def get_capital(data):
    country = data["country"]
    response = await openai_client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {
                "role": "user",
                "content": f"What is the capital of {country}? Just name the "
                "city and nothing else",
            },
        ],
    )
    return response.choices[0].message.content.strip()


# Evaluation data
data = [
    {"data": {"country": "Canada"}, "target": {"capital": "Ottawa"}},
    {"data": {"country": "Germany"}, "target": {"capital": "Berlin"}},
    {"data": {"country": "Tanzania"}, "target": {"capital": "Dodoma"}},
]


def evaluator_A(output, target):
    return 1 if output == target["capital"] else 0


# Create an Evaluation instance
e = Evaluation(
    name="py-evaluation-async",
    data=data,
    executor=get_capital,
    evaluators=[evaluator_A],
    project_api_key=os.environ["LMNR_PROJECT_API_KEY"],
)

# Run the evaluation
asyncio.run(e.run())