umlaut 0.2.2

uMLaut simplifies deploying data science models

uMLaut

The uMLaut library simplifies data science model deployment and querying. It provides a single access point for all your models and an interface to interact with them. uMLaut models can be as extensive as deep learning models or as simple as a reusable code block.

uMLaut offers

• simple commands to track and query models
• history of all model query inputs and results
• model lifecycle management
• access to multiple versions of the same model
• a user interface with MLflow
• model audit tracking history (roadmap)
• auto-deployed models that can be queried through an API (roadmap)

Installing uMLaut

pip install umlaut

---

MLflow Setup

MLflow is a powerful machine learning library created by Databricks for data science teams. It offers an extensive API for tracking and querying models, but the learning curve can be a deterrent for small teams without dedicated data scientists. uMLaut strips away much of the complexity of MLflow while maintaining the immense value of tracking and querying your models in a single location.

MLflow has two requirements:

1. A model artifact storage location

• This can be a local directory or a cloud storage URI. More info in the MLflow docs.

2. A model registry

• The model registry is where model changes and query data are stored. More info in the MLflow docs.

An mlflow server must be running in order to work with uMLaut. The command to start an MLflow server with local artifact storage and a Postgres model registry is as follows:

mlflow server --backend-store-uri postgresql+psycopg2://admin:password@localhost:5432/database --default-artifact-root "mlruns/"

Once the server is running you can navigate to the MLflow UI and begin interacting with models.

---

Core Functionality

uMLaut offers a simple Python class to assist with saving and querying business logic in MLflow.

• track_model: Converts a data science model or block of business logic into an MLflow compatible model
• query_model: Queries a previously trained model and saves audit metadata
• audit_model (roadmap): Retrieve the results of a model run for a historic date

Deploying models with Umlaut

Custom data science models or business logic can be deployed simply by running umlaut.track_model(). Ensure the model code block is in a Python Class and follow the example below.

class ExampleModel():
    """
    Example business logic that can be wrapped into a model.
    The class _must_ contain a 'predict' method.
    """
    def business_logic(self, revenue: int) -> bool:
        return revenue > 5

    def predict(self, model_input: dict) -> bool:
        return self.business_logic(revenue=model_input.get("revenue"))


if __name__ == "__main__":
    """Saves the model to MLflow in an experiment run"""
    from umlaut import Umlaut

    Umlaut.track_model(
        model=ExampleModel(),
        model_name="Quarterly Revenue",
        run_name="Update",
    )

This will push the latest changes of ExampleModel() to MLflow as a new model version. Navigate to the MLflow server where you can find details for the example "Quarterly Revenue" model.

Querying models with Umlaut

Once a model is deployed in MLflow with umlaut.track_model(), it can be queried by calling umlaut.query_model().

from umlaut import Umlaut

result = Umlaut.query_model(
    model_name="Quarterly Revenue",
    input_config={"revenue": 3},
    stage="Staging",
)
print(f"Revenue will{'' if result else ' not'} exceed target")

If we query the simple Quarterly Revenue example model with revenue = 3, the model will return False as the revenue does not exceed the target of 5. The call to the model will be tracked in MLflow with model inputs and results.