Paligemma2 10b Mix 224

🚀 PaliGemma 2 model card

PaliGemma 2 is an advanced vision - language model that combines the capabilities of Gemma 2. It can take both images and text as input and generate text outputs, supporting multiple languages. This model is fine - tuned on a variety of academic tasks and is suitable for a wide range of vision - language applications.

🚀 Quick Start

Access PaliGemma on Hugging Face

To access PaliGemma on Hugging Face, you’re required to review and agree to Google’s usage license. To do this, please ensure you’re logged - in to Hugging Face and click below. Requests are processed immediately. Acknowledge license

Using the Model

You can use the following prompt templates to perform different tasks:

• "cap {lang}": Raw short caption (from WebLI - alt)
• "caption {lang}": Nice, COCO - like short captions
• "describe {lang}": Longer, more descriptive captions
• "ocr": Optical character recognition
• "answer {lang} {question}": Question answering about the image contents
• "question {lang} {answer}": Question generation for a given answer
• "detect {object} ; {object}": Locate listed objects in an image and return the bounding boxes for those objects
• "segment {object}": Locate the area occupied by the object in an image to create an image segmentation for that object

from transformers import (
    PaliGemmaProcessor,
    PaliGemmaForConditionalGeneration,
)
from transformers.image_utils import load_image
import torch

model_id = "google/paligemma2-10b-mix-224"

url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg"
image = load_image(url)

model = PaliGemmaForConditionalGeneration.from_pretrained(model_id, torch_dtype=torch.bfloat16, device_map="auto").eval()
processor = PaliGemmaProcessor.from_pretrained(model_id)

prompt = "describe en"
model_inputs = processor(text=prompt, images=image, return_tensors="pt").to(torch.bfloat16).to(model.device)
input_len = model_inputs["input_ids"].shape[-1]

with torch.inference_mode():
    generation = model.generate(**model_inputs, max_new_tokens=100, do_sample=False)
    generation = generation[0][input_len:]
    decoded = processor.decode(generation, skip_special_tokens=True)
    print(decoded)

Here is a notebook that showcases fine - tuning PaliGemma 2.

✨ Features

• Multimodal Input: Accepts both images and text as input, enabling a wide range of vision - language tasks.
• Multiple Languages Support: Capable of generating text in multiple languages, suitable for global applications.
• Versatile Task Handling: Can perform tasks such as short and long captioning, optical character recognition, question answering, object detection and segmentation.

📦 Installation

The model is available in the transformers library. You can install transformers using the following command:

pip install transformers

📚 Documentation

Model information

Model summary

PaliGemma 2 is an update of the PaliGemma vision - language model (VLM) which incorporates the capabilities of the Gemma 2 models. The PaliGemma family of models is inspired by PaLI - 3 and based on open components such as the SigLIP vision model and Gemma 2 language models. It takes both image and text as input and generates text as output, supporting multiple languages. It is designed for class - leading fine - tune performance on a wide range of vision - language tasks such as image and short video caption, visual question answering, text reading, object detection and object segmentation.

Model architecture

PaliGemma 2 is the composition of a Transformer decoder and a Vision Transformer image encoder. The text decoder is initialized from Gemma 2 in the 2B, 9B, and 27B parameter sizes. The image encoder is initialized from SigLIP - So400m/14. Similar to the original PaliGemma model, PaliGemma 2 is trained following the PaLI - 3 recipes.

Inputs and outputs

• Input: Image and text string, such as a prompt to caption the image, or a question.
• Output: Generated text in response to the input, such as a caption of the image, an answer to a question, a list of object bounding box coordinates, or segmentation codewords.

Model data

Pre - train datasets

PaliGemma 2 is pre - trained on the following mixture of datasets:

• WebLI: WebLI (Web Language Image) is a web - scale multilingual image - text dataset built from the public web. A wide range of WebLI splits are used to acquire versatile model capabilities, such as visual semantic understanding, object localization, visually - situated text understanding, and multilinguality.
• CC3M - 35L: Curated English image - alt_text pairs from webpages ([Sharma et al., 2018](https://aclanthology.org/P18 - 1238/)). We used the Google Cloud Translation API to translate into 34 additional languages.
• VQ²A - CC3M - 35L/VQG - CC3M - 35L: A subset of VQ2A - CC3M ([Changpinyo et al., 2022a](https://aclanthology.org/2022.naacl - main.142/)), translated into the same additional 34 languages as CC3M - 35L, using the Google Cloud Translation API.
• OpenImages: Detection and object - aware questions and answers (Piergiovanni et al. 2022) generated by handcrafted rules on the OpenImages dataset.
• WIT: Images and texts collected from Wikipedia (Srinivasan et al., 2021).

PaliGemma 2 is based on Gemma 2, and you can find information on the pre - training datasets for Gemma 2 in the Gemma 2 model card.

Data responsibility filtering

The following filters are applied to WebLI, with the goal of training PaliGemma 2 on safe and responsible data:

• Pornographic image filtering: This filter removes images deemed to be of pornographic nature.
• Text safety filtering: We identify and filter out images that are paired with unsafe text. Unsafe text is any text deemed to contain or be about child sexual abuse imagery (CSAI), pornography, vulgarities, or is otherwise offensive.
• Text toxicity filtering: We further use the Perspective API to identify and filter out images that are paired with text deemed insulting, obscene, hateful or otherwise toxic.
• Text personal information filtering: We filtered certain personal information and other sensitive data using the Cloud Data Loss Prevention (DLP) API to protect the privacy of individuals. Identifiers such as social security numbers and [other sensitive information types](https://cloud.google.com/sensitive - data - protection/docs/high - sensitivity - infotypes - reference?_gl=1jg604m_gaODk5MzA3ODQyLjE3MTAzMzQ3NTk._ga_WH2QY8WWF5*MTcxMDUxNTkxMS4yLjEuMTcxMDUxNjA2NC4wLjAuMA..&_ga=2.172110058.-899307842.1710334759) were removed.
• Additional methods: Filtering based on content quality and safety in line with our policies and practices.

Implementation information

Hardware

PaliGemma 2 was trained using the latest generation of Tensor Processing Unit (TPU) hardware (TPUv5e).

Software

Training was completed using JAX, Flax, TFDS and [big_vision](https://github.com/google - research/big_vision).

JAX allows researchers to take advantage of the latest generation of hardware, including TPUs, for faster and more efficient training of large models.

TFDS is used to access datasets and Flax is used for model architecture. The PaliGemma 2 fine - tune code and inference code are released in the big_vision GitHub repository.

Evaluation information

Benchmark results

In order to verify the transferability of PaliGemma 2 to a wide variety of academic tasks, we fine - tune the pretrained models on each task. We report results on different resolutions to provide an impression of which tasks benefit from increased resolution and which tasks benefit from larger models. Importantly, none of these tasks or datasets are part of the pretraining data mixture, and their images are explicitly removed from the web - scale pre - training data.

🔧 Technical Details

Model Architecture

PaliGemma 2 combines a Transformer decoder and a Vision Transformer image encoder. The text decoder is initialized from Gemma 2 with different parameter sizes (2B, 9B, and 27B). The image encoder is initialized from SigLIP - So400m/14. The training process follows the PaLI - 3 recipes.

Training Data and Filtering

The model is pre - trained on a mixture of datasets, including WebLI, CC3M - 35L, VQ²A - CC3M - 35L/VQG - CC3M - 35L, OpenImages, and WIT. To ensure the safety and responsibility of the training data, multiple filters are applied to WebLI, such as pornographic image filtering, text safety filtering, text toxicity filtering, and text personal information filtering.

📄 License

The model is released under the Gemma license.

Model page: PaliGemma Resources and technical documentation:

• [PaliGemma 2 on Kaggle](https://www.kaggle.com/models/google/paligemma - 2)
• Responsible Generative AI Toolkit Terms of Use: Terms Authors: Google