harshavaishnav/

DPO_Dataset_2

	Dataset Card for DPO_Dataset_2






	


	Dataset Details

This dataset named DPO_Dataset_2 is made for Direct Preference Optimization of LLM that is finetuned or a base model.

The dataset is made from this playpen-paper-2025/DPO/dpo_dataset_creator.py file.These are the dialogues created from the file for doing a task.The ratio of the positive to negative in the dataset is 1:2.

Curated by: Velineni Harshavaishnav,Vasamsetti Nihal Tej… See the full description on the dataset page: https://huggingface.co/datasets/harshavaishnav/DPO_Dataset_2.