mirror of
https://github.com/MLSysBook/TinyTorch.git
synced 2026-06-03 18:50:52 -05:00
beccbae2ef
Complete capstone competition implementation: - Two division tracks: Closed (optimize) and Open (innovate) - Baseline CNN model for CIFAR-10 - Validation and submission generation system - Integration with Module 19 normalized scoring - Honor code and GitHub repo submission workflow - Worked examples and student templates Module 20 is now a pedagogically sound capstone that applies all Optimization Tier techniques in a fair competition format.
60 lines
1.9 KiB
YAML
60 lines
1.9 KiB
YAML
name: "Competition & Validation"
|
|
module_number: "20"
|
|
description: "TorchPerf Olympics preparation - validation, baseline, and competition submission"
|
|
difficulty: "⭐⭐⭐" # 3 stars - capstone integration
|
|
estimated_time: "1-2 hours"
|
|
|
|
prerequisites:
|
|
- "Module 19: Benchmarking"
|
|
- "Modules 14-18: Optimization techniques"
|
|
|
|
learning_objectives:
|
|
- "Validate TinyTorch installation and environment"
|
|
- "Generate baseline performance metrics"
|
|
- "Understand complete optimization workflow"
|
|
- "Create standardized competition submissions"
|
|
|
|
key_concepts:
|
|
- "System validation and environment checks"
|
|
- "Baseline generation and reference metrics"
|
|
- "End-to-end optimization workflow"
|
|
- "Competition submission format"
|
|
|
|
skills_developed:
|
|
- "Systematic validation and testing"
|
|
- "Performance measurement and comparison"
|
|
- "Integration of multiple optimization techniques"
|
|
- "Professional submission preparation"
|
|
|
|
exports_to: "tinytorch/competition/submit.py"
|
|
|
|
test_coverage:
|
|
- "Installation validation"
|
|
- "Baseline generation"
|
|
- "Worked example workflow"
|
|
- "Competition template structure"
|
|
|
|
connections:
|
|
builds_on:
|
|
- "Module 19 for benchmarking tools"
|
|
- "Modules 14-18 for optimization techniques"
|
|
enables:
|
|
- "TorchPerf Olympics competition participation"
|
|
- "Systematic performance optimization"
|
|
- "Professional ML systems workflow"
|
|
|
|
notes: |
|
|
This is the capstone module that brings together all previous modules.
|
|
It's lightweight (no new techniques) but shows the complete workflow from
|
|
validation through optimization to submission.
|
|
|
|
Students learn:
|
|
1. How to validate their environment works
|
|
2. What baseline performance looks like
|
|
3. How to apply optimizations systematically
|
|
4. How to package work for competition
|
|
|
|
The module includes a complete worked example and a template for students
|
|
to implement their own optimization strategies.
|
|
|