metadata
license: apache-2.0
datasets:
- PKU-Alignment/PKU-SafeRLHF
language:
- en
base_model:
- Qwen/Qwen2-VL-7B-Instruct
pipeline_tag: visual-question-answering
tags:
- Safety
- MLLM
Qwen2-VL-7B finetuned on a mini subset of SafeRLHF along with additional responses.
for more detailed about training data and parameters, please refer to our Paper