LavBench Documentation
Contents:
- Introduction
- LavBench System Architecture & Infrastructure
- 1. System Overview
- 2. Component Technology Stack
- 3. Authentication & Authorization Flow
- 4. Submission Pipeline & Sandbox Isolation Security
- 5. Task Image Build Pipeline & Build Error Taxonomy
- 6. Evaluation Engine Architecture
- 7. API Type Pipeline & Validation Architecture
- 8. SSE Real-Time Streaming Architecture
- 9. Automated & Manual Backup Retention Architecture
- 10. Worker Hardware Budgeting & Clamping
- Competitor Complete Guide
- Table of Contents
- 1. Getting Started and Interface Navigation
- 2. Stage and Task Overview
- 3. Baseline Notebooks and Output Schema
- 4. Notebook Submission Workflow and AST Pre-Validation
- 5. Submission Pipeline and Live Streaming Logs
- 6. Leaderboard, Double-Blind Privacy and Final Selection
- 7. Troubleshooting and Error Matrix
- Jury Portal Complete Guide
- Table of Contents
- 1. Jury Role and Permission Matrix
- 2. Challenge Assignment and JuryChallenge Mapping
- 3. Competitor Onboarding and One-Time Credential Slips
- 4. Live Submission Tracking and Baseline Verification
- 5. Task Environment Build Error Diagnostics
- 6. Custom Evaluator Scoring and Leaderboard Inspection
- 7. Manual Scoring and Score Justification Audit Trail
- 8. Competitor Disqualification and Data Exports
- Administrator Portal Complete Guide
- Initial Setup
- Table of Contents
- 1. Challenge and Stage Lifecycle Management
- 2. Sandbox Customization and Image Build Error Remediation
- 3. Hugging Face Pre-Fetching and Offline Assets
- 4. AST Code Security and Dynamic Metrics Engine
- 5. Custom Evaluator Scripts Logic and Baseline Verification
- 6. User Management, CSV Imports and One-Time Credentials
- 7. Automated and Manual Backup Retention Rules
- 8. Worker Authentication, CLI Setup and SSE Telemetry
- 9. Audit Logging and Score Justification Trail
- API Reference
- API Specification