eval_752
  • English

      • User Guide

      This guide covers the tasks you'll do repeatedly in eval_752: setting up providers, importing datasets, running evaluations, and comparing results.

      If you're new, work through these in order — each step builds on the previous one:

      1. Providers — Connect your API endpoints and verify they work
      2. Datasets — Import or build the question sets you'll evaluate against
      3. Running Evaluations — Launch runs and monitor progress
      4. Viewing Results — Understand what the Dashboard, Runs, and Comparison pages show
      5. Exporting Results — Save and share your evidence

      Once the core workflow feels solid, explore:

      1. Browser Harness — Evaluate models that only exist behind a web UI
      2. Scheduled Evaluations — Automate recurring checks
      3. Advanced Features — Judge scoring, variants, and what's coming next
      Tip

      Don't jump into schedules or Browser Harness until you can run and compare a normal evaluation end-to-end. Debugging too many moving parts at once is frustrating.