EssayEvaluator: Automated Essay Scoring
State-of-the-Art Automated Essay Scoring system using a 3.8B parameter language model that outperforms larger commercial models.
See on GitHub“My dissertation, proving small models can beat giants when tuned properly. (A+ with Distinction)”
Problem
Commercial essay scoring systems are expensive and require large models. Smaller, affordable alternatives sacrifice accuracy significantly.
Solution
Fine-tuned a 3.8B parameter model that achieves QWK of 0.82, reducing error rates by 65% over baselines while costing 99.99% less than commercial alternatives.
Overview
This AI model provides accurate scoring for academic essays, matching industry standards with a QWK of 0.82. The system reduces error rates by nearly 65% compared to baseline models while offering a 99.99% cost reduction compared to commercial alternatives.
Ask my agent about EssayEvaluator
It knows the problem, the build and Daniel's part in it, and can book a call if you want to go deeper.
Ask the agent