PyTorch Lightning has released a new EcoSystem CI project, a lightweight repository that provides easy configuration of ‘Continuous Integration’ running on CPU and GPU. One of the main goals of Ecosystem CI is to enable early discovery of issues through regular testing against stable and development versions of Lightning.
This new integration platform for Data Science and Deep Learning packages based on Lightning boosts compatibility, performs regular testing, and raises early warnings in case of possible collisions. Any user who wants to keep their project aligned with current and future Lightning releases can use the EcoSystem CI to configure their integrations.
Jirka Borovec, Sr Research Engineer at Grid.ai, PyTorch Lightning, wrote in a blog post introducing the project, “We designed the EcoSystem CI to provide a unified structure flexible enough to cover all practical integration needs. The integrations leverage GitHub actions (CPU with many OS versions) and Azure pipelines (GPU with Linux only). We also provide native parallelization, so all projects are tested concurrently using caching to speed-up environment creation.”
The use of Ecosystem CI directly provides out-of-the-box nightly testing on CPU and multi-GPU compute. The project can also be forked and run with your own custom environments and compute resources.
The platform runs two procedures – Prepare Environment and Install all Dependencies, and Copy and Execute all Linked Integration Tests. These steps can be easily extended and used with any other CI system such as CircleCI if testing is required on different types of computing.