---
base_model:
- Xiaodong/Next-DPO-iter1
datasets:
- Xiaodong/DPO-iter2-data-8k
---

wandb task:
https://wandb.ai/xiaodongwang/llava-next-jf-4A100/runs/ck4jmn3r/overview

Dataset:
llava-hound QA
https://huggingface.co/Xiaodong/Next-DPO-iter2/resolve/main/aug_f4_add_chosen_0_8000.jsonl