MilchstraB / LVLM-OnPolicy Public

The official code of paper "Beyond Reward Models: A Classification Approach for Hallucination Mitigation of LVLMs"

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 19 Commits
eval		eval
scripts		scripts
README.md		README.md
data_processor.py		data_processor.py
get_preferdata_cot.py		get_preferdata_cot.py
hallucinating_rewrite.py		hallucinating_rewrite.py
inference.py		inference.py
requirements.txt		requirements.txt
train.py		train.py
utils.py		utils.py