AimKP Codebase Augmenting Intra-Modal Understanding in MLLMs for Robust Multimodal Keyphrase Generation Install Install Package conda create -n llava python=3.10 -y conda activate llava pip install --upgrade pip # enable PEP 660 support pip install -e . pip install -e ".[train]" pip install -r requirements.txt Training Image data of CMKP dataset can be found in CMKP repo MLLM: LLaVA-v1.5-7b, Vision Encoder: clip-vit-large-patch14-336 # For standard training bash /path/to/standard_finetune.sh # For training under AimKP bash /path/to/scripts/AimKP.sh Evaluation python evaluate.py --model-path checkpoint --model-base /path/to/models/llava-v1.5-7b --txt-path "results" Acknowledgement Code is based on LLaVA