Skip to content
Ravindu Weerasinghe
Go back

Small Language Model Fine-Tuning & Evaluation

Researched and implemented fine-tuning strategies such as LoRA, adapters, and prefix tuning.

Improved model performance by ~5% using RAGAS-based evaluation metrics.

Demonstrated feasibility of using local models for privacy-sensitive, domain-specific applications.



Previous Project
Automated Fine-Tuning Pipeline (MLOps)
Next Project
AutoChess – Autonomous Chess Playing Machine