arc/docs/ai-agent/evals/phase-14/README.md
2026-08-21 00:15:57 +00:00

8 lines
250 B
Markdown

# Phase 14 evaluation outputs
- `offline_gold`, `production_control`, and `production_experiment` are separate objects.
- Gold Recall is never Production Recall.
```bash
cd hesabixAPI && .venv/bin/python -m app.services.ai.ai_tool_phase14_eval
```