vermarjun/chad-dpo download history
vermarjun/chad-dpo is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 15 times (3 in the last 7 days), and 61 times in total. It ranks #450,867 among datasets by monthly downloads.
Chad DPO dataset Preference pairs for aligning Chad, the small Hinglish chat model, after supervised fine tuning. DPO shows the model two replies and which one is better, so it learns to pick the sharper, funnier answer on its own. How it was built prompts.jsonl: the u