a2ml - Automation of AutoML

The A2ML ("Automate AutoML") project is a Python API and set of command line tools to automate Automated Machine Learning tools from multiple vendors. The intention is to provide a common API for all Cloud-oriented AutoML vendors. Data scientists can then train their datasets against multiple AutoML models to get the best possible predictive model. May the best "algorithm/hyperparameter search" win. Full documentation for A2ML is available at a2ml.org

The PREDIT Pipeline

Every AutoML vendor has their own API to manage the datasets and create and manage predictive models. They are similar but not identical APIs. But they share a common set of stages:

• Importing data for training
• Train models with multiple algorithms and hyperparameters
• Evaluate model performance and choose one or more for deployment
• Deploy selected models
• Predict results with new data against deployed models
• Review performance of deployed models

Since ITEDPR is hard to remember we refer to this pipeline by its conveniently mnemonic anagram: "PREDIT" (French for "predict"). The A2ML project provides classes which implement this pipeline for various Cloud AutoML providers and a command line interface that invokes stages of the pipeline.

Setup

A2ML is distributed as a python package, so to install it:

$ pip install -U a2ml

It will install Auger provider.

To use Azure AutoML:

Mac:

$ brew install libomp

For Mac OS High Sierra and below:

$ SKLEARN_NO_OPENMP=1 pip install "scikit-learn==0.21.3"
$ pip install "a2ml[azure]" --ignore-installed onnxruntime onnx nimbusml

Linux:

$ apt-get update && apt-get -y install gcc g++ libgomp1

$ pip install "a2ml[azure]"

To use Google Cloud:

$ pip install "a2ml[google]"

To install everything including testing and server code:

$ pip install "a2ml[all]"

Development

To release a new version the flow should be:

1. Change the __version__ variable in a2ml/__init__.py to match what you want to release, minus the "v". By default it would be ".dev0", for example "0.3.0.dev0". This ensures we don’t accidentally release a dev version to pypi.org. So for when we’re ready to release 0.3.0, the __version__ variable should simply be "0.3.0".

2. Commit and push the changes above.

git tag v<the-version> (for example: git tag v0.3.0)
git push --tags

3. verify circleci build passed and docker image tag exists:

pip install -U a2ml==0.3.0
docker pull augerai/a2ml:v0.3.0

4. Increment the __version__ variable in a2ml/__init__.py to the next version in the current milestone. For example, "0.3.1.dev0"