Modelpublic

gokaygokay/paligemma-rich-captions

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face

8likes25downloads

Model Card

Fine tuned version of PaliGemma model on google/docci dataset with middle size captions between 200 and 350 characters. This model has less halucinations.

pip install git+https://github.com/huggingface/transformers

python

from transformers import AutoProcessor, PaliGemmaForConditionalGeneration
from PIL import Image
import requests
import torch

model_id = "gokaygokay/paligemma-rich-captions"

url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true"
image = Image.open(requests.get(url, stream=True).raw)

model = PaliGemmaForConditionalGeneration.from_pretrained(model_id).to('cuda').eval()
processor = AutoProcessor.from_pretrained(model_id)

## prefix
prompt = "caption en"
model_inputs = processor(text=prompt, images=image, return_tensors="pt").to('cuda')
input_len = model_inputs["input_ids"].shape[-1]

with torch.inference_mode():
    generation = model.generate(**model_inputs, max_new_tokens=256, do_sample=False)
    generation = generation[0][input_len:]
    decoded = processor.decode(generation, skip_special_tokens=True)
    print(decoded)

Academic use and citation

This model is one of the image-captioning systems evaluated in *Automated Interpretation of Non-Destructive Evaluation Contour Maps Using Large Language Models for Bridge Condition Assessment* (IEEE BigData 2024).

If this model is useful in your research, please cite both the model and the paper that documents its academic use:

bibtex

@software{aydogan2024paligemma_rich_captions,
  author = {Aydoğan, Gökay},
  title = {PaliGemma Rich Captions},
  year = {2024},
  url = {https://huggingface.co/gokaygokay/paligemma-rich-captions},
  note = {Hugging Face model}
}

@inproceedings{darji2024automated,
  author = {Darji, Viraj Nishesh and Liao, Callie C. and Liao, Duoduo},
  title = {Automated Interpretation of Non-Destructive Evaluation Contour Maps Using Large Language Models for Bridge Condition Assessment},
  booktitle = {2024 IEEE International Conference on Big Data (BigData)},
  year = {2024},
  pages = {3258--3263},
  doi = {10.1109/BigData62323.2024.10825532}
}

Maintainer: Gökay Aydoğan