Installation¶

System requirements¶

Auto-PyTorch has the following system requirements:

Linux operating system (for example Ubuntu) (get Linux here),
Python (>=3.7) (get Python here).
C++ compiler (with C++11 supports) (get GCC here) and
SWIG (version 3.0.* is required; >=4.0.0 is not supported) (get SWIG here).

Installing Auto-Pytorch¶

PyPI Installation¶

Auto-PyTorch for Time Series Forecasting requires additional dependencies

Manual Installation¶

# Following commands assume the user is in a cloned directory of Auto-Pytorch

# We also need to initialize the automl_common repository as follows
# You can find more information about this here:
# https://github.com/automl/automl_common/
git submodule update --init --recursive

# Create the environment
conda create -n autopytorch python=3.8
conda activate autopytorch
conda install swig
cat requirements.txt | xargs -n 1 -L 1 pip install
python setup.py install

Similarly, Auto-PyTorch for time series forecasting requires additional dependencies

Docker Image¶

A Docker image is also provided on dockerhub. To download from dockerhub, use:

docker pull automlorg/autopytorch:master

You can also verify that the image was downloaded via:

docker images  # Verify that the image was downloaded

This image can be used to start an interactive session as follows:

docker run -it automlorg/autopytorch:master

To start a Jupyter notebook, you could instead run e.g.:

docker run -it -v ${PWD}:/opt/nb -p 8888:8888 automlorg/autopytorch:master /bin/bash -c "mkdir -p /opt/nb && jupyter notebook --notebook-dir=/opt/nb --ip='0.0.0.0' --port=8888 --no-browser --allow-root"

Alternatively, it is possible to use the development version of autoPyTorch by replacing all occurences of master by development.