Artanic30
/

DA-DPO_llava_v1.5_7B

Reinforcement Learning

Model card Files Files and versions

Metrics Training metrics Community

This is the official checkpoint released for the TMLR 2025 paper DA-DPO.

The training dataset can be found in BPO.

For model usage, please follow the instructions in LLaVA-v1.5-7B

References

Model Paper

Downloads last month: 1

Video Preview

Reinforcement Learning

loading

Model tree for Artanic30/DA-DPO_llava_v1.5_7B

Base model

liuhaotian/llava-v1.5-7b

Finetuned

(26)

this model

Collection including Artanic30/DA-DPO_llava_v1.5_7B

DA-DPO

This is the collection for the TMLR 25 paper DA-DPO. Project Page: https://artanic30.github.io/project_pages/DA-DPO/ • 3 items • Updated Jan 25 • 1

Paper for Artanic30/DA-DPO_llava_v1.5_7B

DA-DPO: Cost-efficient Difficulty-aware Preference Optimization for Reducing MLLM Hallucinations

Paper • 2601.00623 • Published Jan 2