Best practices
Stratix best practices — quick reference for new users.
Last updated
Was this helpful?
⌘Ctrlk
Stratix best practices — quick reference for new users.
Pick 3-5 benchmarks that match your task; resist "more"
Re-evaluate quarterly as the model frontier moves
Use spaces for stable configs; one-offs for exploration
GEPA-optimize against ≥30 labels
Hold out a validation set
Watch for over-fit (validation agreement << training)
Sample production traffic; full corpus is rarely worth it
Anchor with deterministic rules; reserve judges for residual subjective bar
Wire CI gates after you've trusted the eval interactively
Per-feature tolerances — not every feature has the same bar
Last updated
Was this helpful?
Was this helpful?