Warrior User Guide

The documents in this section provide instructions for setting up specific functionality within Warrior. To learn about the underlying algorithms used in various parts of the product, please see Algorithms.

• Algorithmic Bias
  • Monitoring Attributes for Bias
    • Model Input Attributes
    • Non-Input Attributes
    • Viewing Fairness Metrics
  • Mitigation Methods
• Explainability
  • Setting up your project directory
    • Requirements File
    • Prediction Function
    • Load Image Function
    • Project Structure
  • Enabling Explainability
  • Troubleshooting
    • AttributeError When Loading Predict Function
• Enrichments
  • General Usage
  • Enrichment List
    • Anomaly Detection
    • Bias Mitigation
    • Explainability
    • Hotspots
• Integrations
  • SparkML Integration
    • Train and Save SparkML Model
    • Onboard to Warrior
    • Set reference data
    • Enabling Explainability
    • Send a batch of inferences
  • AWS SageMaker Data Capture Integration
    • Prerequisites
    • SageMaker Configuration
      • Configuring Data Capture through the SageMaker SDK
    • AWS Lambda Setup
      • Lambda Function Configurations
      • Lambda Function Trigger
      • Lambda Code
    • Summary
  • Batch Ingestion from S3
    • Obtain Credentials
    • Create Prefix
      • Inferences
      • Ground Truth
      • Prefix Parameters
    • Upload Inference Files
    • Upload Ground Truth Files