User Guide

This section explains how to configure KKT-HardNet, define constrained problems, attach data, train a model, reload it, and run inference.

• Configuration Options
  • Example
• Defining the Problem
  • Creating the Model Object
  • Defining Parameters and Variables
  • Inverse Parameters
  • Defining the Objective
  • Using Extracted Matrices
  • Defining Constraints
  • Bounds
• Defining Parameter Data
  • CSV-based Dataset
  • Data Requirements by Workflow
  • Example
  • Notes
• Training and Inference
  • Build and Train
  • Model Summary and Training History
  • Saved Artifacts
  • Reloading the Model
  • Projection Backends