Home / Projects / mlops-experiment-harness
ACTIVEPOD 06: Systems, HPC & MLOps

Automated Experiment Lineage & GPU Telemetry Testbed

Project LeadArjun S.
CycleCycle 2026.1
Reproducibility StatusPublic Code Available

Objective & Abstract

Building lightweight GitHub Actions runners that automatically capture VRAM spikes, seed integrity, and model hash lineage per pull request.

Investigated Methodologies

  • Containerized runner sandboxing
  • Prometheus CUDA exporter scrape
  • SHA-256 artifact signing

Datasets & Benchmarking Corpora

📊Standard Lab Benchmark Runner Suites

All training and evaluation datasets undergo ethics audit and licensing compliance per AIAC Research Council standards.

Associated Technical Output

open source

AIAC MLOps Experiment Runner Template & GitHub Actions Suite

Arjun S., Karthik P.

Turnkey CI/CD GitHub Actions workflow for machine learning repositories that captures GPU VRAM usage via Prometheus exporters, validates deterministic random seeds, and computes dataset SHA-256 hashes on every commit.