Update README.md

README.md

@@ -195,7 +195,7 @@ Dataset: [HuggingFaceH4/ultrafeedback_binarized](https://huggingface.co/datasets
 <br><br>
 
 ## Training Procedure
-Both Pre-Train and Finetuning used [our fork](https://github.com/Pints-AI/1.5-Pints) of the [LitGPT Framework](https://github.com/Lightning-AI/litgpt). For DPO, we used the methods set out in [The Alignment Handbook](https://github.com/huggingface/alignment-handbook/blob/main/scripts/run_dpo.py). More details can be found in our [paper](TOBEADDED).
+Both Pre-Train and Finetuning used [our fork](https://github.com/Pints-AI/1.5-Pints) of the [LitGPT Framework](https://github.com/Lightning-AI/litgpt). For DPO, we used the methods set out in [The Alignment Handbook](https://github.com/huggingface/alignment-handbook/blob/main/scripts/run_dpo.py). More details can be found in our [paper](https://arxiv.org/abs/2408.03506).
 
 ## Training Hyperparameters
 **Pre-Train**<br>