Documentation

Welcome to SigOpt’s developer documentation. If you have a question you can’t answer, feel free to contact us!
This feature is currently in Beta. You can request free access on the main page of the beta or contact us directly for more information.

Get Started

If you haven't gotten an overview yet, take a quick look at the basic concepts.

There are two steps to getting set up.

  1. Installation
  2. First Time Setup & Configuration

1. Installation

Note: If you have previously installed SigOpt's Orchestrate package, please uninstall it first: pip uninstall orchestrate.

Install the SigOpt Python package:

pip install --upgrade https://sigopt-python-mpm.s3-us-west-1.amazonaws.com/sigopt-7.1.4-py2.py3-none-any.whl

That's all! You should be able to run sigopt --help.

2. First Time Setup & Configuration

In order to interact with the SigOpt API, you'll need to set up some general, basic configuration information.

Configure the SigOpt CLI with the command sigopt config. You'll be asked a short set of questions:

$ sigopt config
SigOpt API token (find at https://app.sigopt.com/tokens/info):

Log Collection
        This will capture and upload the standard output and standard error of your
        Runs so that you can view them on the SigOpt dashboard.
Enable log collection [y/N]:

Code Tracking
        This will record and upload the content of your code so that you can view it
        on the SigOpt dashboard.
        The source code hash from your VCS (ie. Git) will be recorded even if this
        option is disabled.
Enable code tracking [y/N]:

You may override the config by setting the SIGOPT_API_TOKEN environment variable. Visit the authentication page for more details about tokens.

Next: Instrument Your Training Runs

The next step is to instrument your training runs with SigOpt's API.

Choose your development environment:

Or convert an existing experiment.