Hyperpod Version Checker

If you're running SageMaker HyperPod clusters, you know version mismatches across nodes can kill a training run. This script checks everything that matters: NVIDIA drivers, CUDA, cuDNN, NCCL, EFA, MPI, Neuron SDK for Trainium, and the usual Python/PyTorch stack. It's a bash script you upload via the hyperpod-ssm skill and run on each node, with JSON output that makes diffing across nodes straightforward. The real value is the compatibility analysis built in, like catching when your driver doesn't support your CUDA version or when NCCL versions aren't consistent. Saves you from debugging distributed training failures that turn out to be a version mismatch on node 7.

skills.sh installs: 149
GitHub stars: 938
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add awslabs/agent-plugins --skill hyperpod-version-checker --agent claude-code

Installs into .claude/skills of the current project.

Files

HyperPod Version Checker

Upload to cluster nodes via hyperpod-ssm skill, then execute.

Usage

# Text report to console + file
bash hyperpod_check_versions.sh

# JSON only to stdout (text report still saved to file) — best for piping/parsing
bash hyperpod_check_versions.sh --json

# Custom output file
bash hyperpod_check_versions.sh --output /tmp/versions.txt

# No color (for logging)
bash hyperpod_check_versions.sh --no-color

Output file: component_versions_<hostname>_<timestamp>.txt (default)