Exporting models (building)

Saving and loading models is fun, but the real value of a model is how you can use it. To make it easier to get models to production, MLEM has 3 related functionalities: building, serving, and deploying. We’ll start with building.

Building is a way to “bake” your model into something usable in production like a Docker image, or export your model into another format. For this tutorial we will create a pip-ready package from our model. You can see the full list of available builders here.

Creating Python package

To create a build/ directory with pip package run this command:

$ mlem build rf pip -c target=build/ -c package_name=example_mlem_get_started
⏳️ Loading model from .mlem/model/rf.mlem
💼 Written `example_mlem_get_started` package data to `build`

In this command, we specified that we want to build rf model with pip builder and then provided two arguments, target is the directory where the builder will write all the files and package_name is the name of our package.

Exploring Python package

Let’s see what we’ve got

$ tree build/
build/
├── MANIFEST.in
├── example_mlem_get_started
│   ├── __init__.py
│   ├── model
│   └── model.mlem
├── requirements.txt
└── setup.py

As you can see, the builder generated all the files necessary for a python package. This includes sources, requirements, setup.py, and the model itself.

Using Python package

Now you can distribute and install the package. Its code declares all the same methods our model had, so you can try to use it like this:

import example_mlem_get_started

example_mlem_get_started.predict(df)

Pre-configured builders

Alternatively, you can pre configure your builder in the form of yaml file either manually or via mlem declare command which uses the same interface with multiple -c options like this:

$ mlem declare builder pip pip_config \
  -c target=build/ -c package_name=example_mlem_get_started
💾 Saving builder to .mlem/builder/pip_config.mlem
$ cat .mlem/builder/pip_config.mlem
object_type: builder
package_name: example_mlem_get_started
target: build/
type: pip

Now you can use this config as a value for --load option in mlem build

$ mlem build rf -l pip_config
⏳️ Loading builder from .mlem/builder/pip_config.mlem
⏳️ Loading model from .mlem/model/rf.mlem
💼 Written `example_mlem_get_started` package data to `build`

$ git add .mlem/packager/pip_config.mlem
$ git commit -m "Add package config"
$ git diff 4-pack

Also, you can do all of this programmatically via Python API:

from mlem.api import build, load_meta

build("pip", "rf", target="build", package_name="example_mlem_get_started")
build(load_meta("pip_config"), "rf")