SmolLM-135M: Base vs SFT vs SFT+DPO

Compare outputs across the full fine-tuning pipeline. Built for IBA NLP Assignment 04.

Examples