pritamqu's picture
Update README.md
bfcbb5a verified
metadata
license: apache-2.0
base_model:
  - OpenGVLab/VideoChat2_stage3_Mistral_7B

Installation

Clone the repository and navigate to the RRPO directory:

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

conda create -n videochat2 python=3.10 -y
conda activate videochat2
pip install -r videochat2.txt

Download weights

# base model
git clone [email protected]:OpenGVLab/VideoChat2_stage3_Mistral_7B
# RRPO weights
git clone [email protected]:pritamqu/VideoChat2_stage3_Mistral_7B-RRPO-16f-LORA

Inference

conda activate videochat2
BASE_WEIGHTS="./VideoChat2_stage3_Mistral_7B"
WEIGHTS_ROOT="./"

python inference.py \
    --base_model_name "videochat2_mistral_7b" \
    --model-path ${BASE_WEIGHTS} \
    --model-path2 ${WEIGHTS_ROOT}"/VideoChat2_stage3_Mistral_7B-RRPO-16f-LORA" \
    --video_path "sample_video.mp4" \
    --question "Describe this video." \
    --model_max_length 1024