MLFlow Lab Runbook 1) Start Infra cd ~/.openclaw/workspace/projects/mlflow-lab/infra cp .env.example .env docker compose up -d 2) Validate Services MLflow UI/API: http://localhost:5000 MinIO Console: http://localhost:9001 Postgres: localhost:5433 3) Client Setup export MLFLOW_TRACKING_URI=http://127.0.0.1:5000 4) First Experiment (PyTorch/HF) Run a baseline script in services/training/. Log params, metrics, and model artifacts. 5) Promotion Policy Register model version after successful run. Move to Staging only if evaluation thresholds pass. Move to Production after manual approval. 6) Backups Snapshot Postgres metadata Snapshot MinIO bucket (artifacts)