---
license: apache-2.0
base_model:
- lmms-lab/LLaVA-Video-7B-Qwen2
---

<a href='https://arxiv.org/abs/2504.12083'><img src='https://img.shields.io/badge/arXiv-paper-red'></a> 
<a href='https://pritamqu.github.io/RRPO/'><img src='https://img.shields.io/badge/project-RRPO-blue'></a> 
<a href='https://huggingface.co/datasets/pritamqu/self-alignment'><img src='https://img.shields.io/badge/huggingface-datasets-green'></a> 
<a href='https://huggingface.co/collections/pritamqu/rrpo-67fbc8c048b298a5fdfb167b'><img src='https://img.shields.io/badge/model-checkpoints-yellow'></a> 
</a><a href='https://github.com/pritamqu/RRPO'><img src='https://img.shields.io/badge/github-repository-purple'></a> 


## Installation

Clone the repository and navigate to the RRPO directory:

```sh
git clone https://github.com/pritamqu/RRPO
cd RRPO

conda create -n llava python=3.10 -y
conda activate llava
pip install -r llavavideo.txt
```

## Download weights

```
# base model
git clone git@hf.co:lmms-lab/LLaVA-Video-7B-Qwen2
# RRPO weights
git clone git@hf.co:pritamqu/LLaVA-Video-7B-Qwen2-RRPO-32f-LORA
```

## Inference

```
conda activate llava
BASE_WEIGHTS="./LLaVA-Video-7B-Qwen2"
WEIGHTS_ROOT="./"

## using lora weights
python inference.py \
    --base_model_name "llavavideo_qwen_7b" \
    --model-path ${BASE_WEIGHTS} \
    --model-path2 ${WEIGHTS_ROOT}"/LLaVA-Video-7B-Qwen2-RRPO-32f-LORA" \
    --video_path "sample_video.mp4" \
    --question "Describe this video." \
    --model_max_length 1024

```