OpenXLA is a community-driven and modular open source compiler. It will enable efficient lowering, optimization and deployment of ML models from most major frameworks to any hardware backend notably CPUs, GPUs, and ML ASICs.

It is currently in the process of being created from the code currently inside tensorflow, under the OpenXLA SIG governance.

• For technical questions, contact Mehdi Amini - aminim at google
• For administrative questions, contact Thea Lamkin - thealamkin at google

• GitHub (current)
• Discord TBA
• Community proposals TBA
• Community meetings TBA

While under TensorFlow governance, all community spaces for SIG OpenXLA are subject to the TensorFlow Code of Conduct.