The PyTorch ci-infra repo contains the terraform configuration for the PyTorch Lambda based EC2 instance autoscaler.

The PyTorch CI infrastructure is for the use of the github.com/pytorch/pytorch project. It powers GitHub Actions self-hosted runners from the PyTorch Foundation's Cloud Provider Accounts.

The CI Infrastructure is widely scoped, and related work is still in progress. As a result, the overall documentation is spread out across various repositories and working groups. A summary of the available learning documentation is provided below.

The PyTorch test-infra project is collection of infrastructure components that are supporting the PyTorch CI/CD system.

Learn about the terraform module that sets up self hosted github runners on AWS along with the infra needed to autoscale them.

If you are interested in contributing to the PyTorch CI providing runners, these guidelines can help you get started.

The test-infra wiki serves to host all of the code used for testing infrastructure across the PyTorch organization.

Learn more about Terraform configurations for managing Datadog monitoring and observability infrastructure for the PyTorch Foundation.

The focus of the PyTorch CI working group is to maintain, improve, monitor, and cost manage the existing PyTorch CI infrastructure.

The PyTorch multi-cloud working group has been created to develop a sustainable, equitable, community managed approach to the CI/CD pipeline for PyTorch in a multi-cloud environment.

This working group complement the work done by the PyTorch CI Working Group by focusing on expanding the CI/CD infrastructure to a multi-cloud environment, in a way that is sustainable for the community and for the CI WG.

You can learn more about proposed architecture, guidelines, and other information here.

To stay uptodated with the ongoing progress, refer to the working group google doc and here.

Learn more about PyTorch HUD and services used by the HUD.

Learn more about the CI/CD security principles.