Skip to content

Conversation

@leo-amd
Copy link
Collaborator

@leo-amd leo-amd commented Nov 19, 2025

Description

This PR introduces a new GitHub Actions workflow (TransformerEngine CI) to orchestrate the build and test process for AMD ROCm platforms.

Motivation:
This workflow ports the existing pipeline from Jenkins CI to GitHub Actions to improve visibility and integration with the repository.

Key Features:

  • Runs on self-hosted runners (linux-mi325-4).
  • Implements manual Docker container management to allow device access (/dev/kfd, /dev/dri) for ROCm.
  • Dynamically selects Docker images based on the target branch (Dev vs Release versions).
  • Executes Build, Single-GPU (sGPU), Multi-GPU (mGPU), and Example tests for PyTorch and JAX.

Copy link
Collaborator

@ipanfilo ipanfilo left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM after CI completion

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants