How mohan-sai-karthikeya101/waypoint works
WayPoint is a student academic-performance analysis system with a separate model-training workflow. Students provide six academic and behavioral inputs; the application predicts a score and learned risk category, explains influential factors, and offers practical recommendations and target-score planning. Training generates and retains data, evaluates fresh and compatible continued-training candidates, then saves versioned models and updates the current models used by the application. The graph follows the README’s documented workflow and the sampled source excerpts; several internal stages are described but not fully shown in source.
Main entry points: training.py Training Pipeline · WayPoint.py Student Application · training Training Batches · testing Testing Batches
Components 19
Model Training
- Training Pipeline
training.py - Data Generation
training.py - Historical Data Loading
training.py - Candidate Training
training.py - Model Evaluation
training.py - Model Selection
training.py
Persistent State
- Training Batches
data/training - Testing Batches
data/testing - Versioned Models
models/versions - Current Models
models/current - Project Registry
project_registry.json
Student Analysis
- Student Application
WayPoint.py - Model Loading
WayPoint.py - Score & Risk Prediction
WayPoint.py - SHAP Explanations
WayPoint.py
Actionable Guidance
- Recommendations
WayPoint.py - Dream Score Planning
WayPoint.py - Analysis Results
WayPoint.py
Other components
- Student
Connections 23
- Training Pipeline → Data Generation: generates batches
README.md - Data Generation → Training Batches: writes batches
README.md - Data Generation → Testing Batches: writes batches
README.md - Training Pipeline → Historical Data Loading: loads history
README.md - Historical Data Loading → Training Batches: reads batches
training.py - Historical Data Loading → Testing Batches: reads batches
training.py - Training Pipeline → Candidate Training: trains candidates
README.md - Candidate Training → Versioned Models: uses prior models
README.md - Training Pipeline → Model Evaluation: evaluates candidates
README.md - Model Evaluation → Model Selection: ranks candidates
README.md - Training Pipeline → Versioned Models: saves new version
README.md - Training Pipeline → Current Models: updates selected models
README.md - Training Pipeline → Project Registry: updates registry
training.py - Student → Student Application: provides inputs
README.md - Student Application → Model Loading: starts analysis
WayPoint.py - Model Loading → Current Models: loads models
README.md - Model Loading → Score & Risk Prediction: supplies models
README.md - Score & Risk Prediction → SHAP Explanations: explains prediction
README.md - Score & Risk Prediction → Recommendations: informs guidance
README.md - Student Application → Dream Score Planning: offers target planning
README.md - Dream Score Planning → Analysis Results: presents plan
README.md - SHAP Explanations → Analysis Results: presents factors
README.md - Recommendations → Analysis Results: presents advice
README.md