Run GitHub Actions on AWS p5.48xlarge

Why this fits

• 192 vCPU and 2048 GiB RAM — sized for workloads the default GitHub-hosted runners can't comfortably host, all on your own AWS account.
• H100 × 8 — for teams that need Hopper-class FP8 in CI.
• NVIDIA H100 80GB for PyTorch / JAX and other CUDA workloads, with drivers + CUDA pre-baked in the runner image.

.cirun.yml

.cirun.yml

1runners:
2  - name: aws-p5-48xlarge
3    cloud: aws
4    instance_type: p5.48xlarge
5    # Use AWS Deep Learning AMI GPU PyTorch on Ubuntu 22.04, or a
6    # Cirun-published H100-compatible AMI.
7    machine_image: ami-04823729c75214919
8    labels:
9      - cirun-aws-p5-48xlarge

Drop this in your repo root. The first workflow that requests the runner label spins this configuration up on your cloud account.