Skip to content

Latest commit

 

History

327 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LCLS Python Environments

A repository holding the definitions of the conda environments used on the SLAC network. Environments are built and published using kubernetes jobs with the appropriate OS images. The environments described by the files are as follows:

python3_rhel7_env

This is the default environment that will be available to anyone logging into a RHEL7 or above box and sourcing the default startup scripts. Intended to be stable, only updated via CATER requests, suitable for running production applications.

rhel7_devel

Updated via a nightly cron job (10 pm) to obtain more recent versions of packages.

rocky9_gpu_devel

An environment with the GPU version of PyTorch for running on CUDA capable rocky 9 systems. Also updated via a nightly cron job (10 pm).

Environment definitions

Each environment is defined by a pixi manifest, pixi.toml, in a directory named after the environment (for example rhel7_devel/pixi.toml). Nightly built ones solve that manifest fresh each night and save the resulting pixi.lock next to the published tarball. Each manifest sets exclude-newer = "7d", so a conda-forge or PyPI release is not picked up until it has been published for at least 7 days. Packages installed from git are not delayed.

Documentation

For additional documentation around how these environments are built, and the nightly environment in particular, see here: https://confluence.slac.stanford.edu/spaces/ppareg/pages/476119856/Nightly+Python+Environment

About

Packed version of python environments

Resources

Stars

3 stars

Watchers

6 watching

Forks

Releases

Packages

Used by

Contributors

Languages