---
license: apache-2.0
---
An experimental DPO finetune of SmartTinyLlama with Alpaca-QLoRA
Datasets
Trained on bagel style dpo datasets
Prompt Template
Uses chatml style prompt template