Claims GPT Evaluation Suite
Evaluation framework for testing and validating Claims GPT models.
Overview
This repository contains evaluation schemas, test cases, and metrics for assessing Claims GPT model performance.
Contents
eval-schema.json - JSON schema for evaluation definitions- Test case templates
- Metric configurations
Usage
- Define test cases following the eval-schema.json format
- Run evaluations against target models
- Review metrics and results
Metrics
- Accuracy threshold
- Latency (ms)
- Cost per evaluation
Organization
Part of GCC Insurance Intelligence Lab Dev environment.
