LLM Evaluators (Configuration)¶
✅ Status: Complete
LLM evaluators are fully implemented as a configuration in Scenario Manager.
Overview¶
LLM evaluators are not a separate application in LLARS. They are configured per scenario and run as automatic evaluators alongside human ratings.
Typical flow: - Enable LLM evaluation and select models in the Scenario Wizard - Evaluation starts automatically after creation (or manually via Scenario Manager) - Progress and results are visible in the Evaluation/Results tabs
Documentation¶
| Document | Description | Status |
|---|---|---|
| Concept | Historical concept from the LLM-as-Judge phase | 🟡 Background |
| Comparison Modes | Comparison modes (Pillar Sample, Round-Robin, Free-for-All) | 🟡 Background |
| Sampling Strategies | Sampling methods for different pillar sizes | 🟡 Background |
Features (current in LLARS)¶
- LLM evaluation as a scenario configuration
- Selection of system and custom provider models
- Optional auto-start after scenario creation
- Live status and results overview in Scenario Manager
Relevant files¶
app/
├── routes/llm/llm_evaluation_routes.py # Start/stop/progress of LLM evaluation
├── routes/scenarios/scenario_manager_api.py # LLM evaluators in scenario config
├── services/llm/llm_ai_task_runner.py # LLM evaluation runner
└── services/evaluation/ # Evaluation logic and aggregation
llars-frontend/src/views/ScenarioManager/
├── components/ScenarioWizard.vue # Configure LLM evaluation
├── components/tabs/ScenarioEvaluationTab.vue # LLM status/results
└── composables/useScenarioManager.js # Start/stop LLM evaluation