Installation

Installation may follow one of three paths: deployment or development for CPU-only execution, or a mixed development and deployment environment with Nvidia CUDA support. Choose the relevant tab for details on each of those installation paths.

Deployment

Install using pip, or your preferred Python project management solution (e.g. uv, rye or poetry).

Stable release, from PyPI:

pip install sleepless

Latest development branch, from its git repository:

pip install git+https://gitlab.idiap.ch/medai/software/sleepless@main

Development

Checkout the repository, and then use pixi to setup a full development environment:

git clone git@gitlab.idiap.ch:medai/software/sleepless
pixi install --frozen
pixi run sleepless --help

Tip

The --frozen flag will ensure that the latest lock-file available with sources is used. If you’d like to update the lock-file to the latest set of compatible dependencies, remove that option.

If you use direnv to setup your pixi environment when you enter the directory containing this package, you can use a .envrc file similar to this:

watch_file pixi.lock
export PIXI_FROZEN="true"
eval "$(pixi shell-hook)"

CUDA

Checkout the repository, and then use pixi to setup a version of this package that can run on a CUDA-enabled machine:

git clone git@gitlab.idiap.ch:medai/software/sleepless
pixi install --frozen -e cuda
pixi run -e cuda sleepless --help

To enable CUDA support, always run applications using the cuda environment via pixi run -e cuda .... Refer to further pixi configuration tips on the Development tab above.

Setup

A configuration file may be useful to setup global options that should be often reused. The location of the configuration file depends on the value of the environment variable $XDG_CONFIG_HOME, but defaults to ~/.config/sleepless.toml. You may edit this file using your preferred editor.

Here is an example configuration file that may be useful as a starting point:

[datadir]
EDF = "/Users/myself/dbs/EDF"
MASS = "/Users/myself/dbs/MASS"

Tip

To get a list of valid data directories that can be configured, execute:

sleepless dataset list

You must procure and download datasets by yourself. The raw data is not included in this package as we are not authorised to redistribute it.

To check whether the downloaded version is consistent with the structure that is expected by this package, run:

sleepless dataset check EDF

Supported Datasets

Here is a list of currently supported datasets in this package, alongside notable properties. Each dataset name is linked to the location where raw data can be downloaded.

Datasets	EEG channels	EOG channels	EMG channels	Annotation method	Train	Validation	Test	Reference
ST-EDF	3	1	1	R&K	28	8	8	[SLEEP_EDF-2018]
SC-EDF	3	1	1	R&K	97	28	28	[SLEEP_EDF-2018]
_SS3-MASS	20	2	3	AASM	33	10	10	[MASS-2014]