How deva1559/validator works
Model Validator AI is an educator-facing platform for deterministic notebook audits. Faculty submit student details and an .ipynb assignment; the backend inspects notebook cells for dataset hygiene, training procedure, leakage, and metric evidence, then scores the run against configurable baselines and stores results for review. The UI also exposes dashboards, student records, a validation queue, leaderboard, and exportable audit reports. The main API excerpt is partial, so some frontend-to-backend links are shown as unverified rather than inferred from matching feature names.
Main entry points: main.py Validation API · App.tsx Application shell · config.ts API configuration · UploadProjects.tsx Project upload
Components 15
Review interface
- Application shell
frontend/src/App.tsx - Navigation
frontend/src/components/Sidebar.tsx - Dashboard
frontend/src/components/Dashboard.tsx
Submission and audit
- Project upload
frontend/src/components/UploadProjects.tsx - API configuration
frontend/src/config.ts - Validation API
backend/main.py - Notebook audit engine
backend/evidence.py
Scoring and records
- Baseline settings
frontend/src/components/BaselineConfig.tsx - Baseline scoring
backend/evidence.py - Validation records
backend/database.py
Audit review
- Validation queue
frontend/src/components/ValidationQueue.tsx - Student records
frontend/src/components/Students.tsx - Leaderboard
frontend/src/components/Leaderboard.tsx - Audit reports
frontend/src/components/Reports.tsx
Other components
- Faculty
Connections 28
- Application shell → Navigation: renders
frontend/src/App.tsx - Application shell → Dashboard: renders
frontend/src/App.tsx - Application shell → Project upload: renders
frontend/src/App.tsx - Application shell → Baseline settings: renders
frontend/src/App.tsx - Application shell → Validation queue: renders
frontend/src/App.tsx - Application shell → Student records: renders
frontend/src/App.tsx - Application shell → Leaderboard: renders
frontend/src/App.tsx - Application shell → Audit reports: renders
frontend/src/App.tsx - Faculty → Project upload: submits notebook
README.md - Project upload → Validation API: uploads assignment
- Validation API → Notebook audit engine: analyzes notebook
backend/main.py - Notebook audit engine → Validation records: writes evidence
backend/evidence.py - Validation API → Validation records: reads and updates
backend/main.py - Baseline settings → Validation API: updates baselines
- Validation API → Baseline scoring: re-evaluates runs
backend/main.py - Baseline scoring → Validation records: stores scores
backend/evidence.py - Validation queue → Validation API: requests leaderboard
- Dashboard → Validation API: loads metrics
- Student records → Validation API: loads student data
- Leaderboard → Validation API: loads rankings
- Audit reports → Validation API: loads audit dossier
- Project upload → API configuration: uses endpoint config
frontend/src/components/UploadProjects.tsx - Validation queue → API configuration: uses endpoint config
frontend/src/components/ValidationQueue.tsx - Student records → API configuration: uses endpoint config
frontend/src/components/Students.tsx - Dashboard → API configuration: uses endpoint config
frontend/src/components/Dashboard.tsx - Leaderboard → API configuration: uses endpoint config
frontend/src/components/Leaderboard.tsx - Audit reports → API configuration: uses endpoint config
frontend/src/components/Reports.tsx - Baseline settings → API configuration: uses endpoint config
frontend/src/components/BaselineConfig.tsx