Script for a deployment, usually score.py: init() loads the model as the container starts and run() handles scoring for each request. MLflow models do not require one.
Also called score.py.
Read more: Microsoft Learn
In the Ultra Transcenders books
Each book explains Scoring script in context, with comparison tables and the common traps.
Terms in this definition
- Model deployment
Inside a resource, each deployment is a named copy of a Foundry model with its own TPM quota and deployment type. Requests identify the model by this deployment name.
- Container
Something that groups data. Blob Storage containers sit inside a storage account and hold blobs much as folders hold files; Cosmos DB containers hold items and set the scope for partitioning and throughput.
- MLflow
An open-source framework for experiment tracking and model management. Azure Machine Learning adopts it as its tracking API, so each workspace acts as a compatible MLflow tracking server.
Related terms
- MLflow model
When a model is stored in MLflow format, with an MLmodel file and conda requirements, you can deploy it to batch or online endpoints with no scoring script or environment of your own.
- No-code deployment
Azure Machine Learning itself supplies the scoring script and environment, so an MLflow model can go to a batch or online endpoint with neither written by you.
- Online deployment
The resources behind an online endpoint, namely model, environment, scoring script, instance type and instance count. One endpoint may contain several, for example blue and green.