google/paligemma2-3b-mix-448

Note

This is a remotely-sourced model from the paligemma2 plugin, maintained by the community. It is not part of FiftyOne core and may have special installation requirements. Please review the plugin documentation and license before use.

PaliGemma 2 mix checkpoints are fine-tuned on a diverse set of tasks and are ready to use out of the box. These tasks include short and long captioning, optical character recognition, question answering, object detection and segmentation, and more..

Details

• Model name: google/paligemma2-3b-mix-448

• Model source: https://huggingface.co/google/paligemma2-3b-mix-448

• Model author: Google DeepMind

• Model license: gemma (https://ai.google.dev/gemma/terms)

• Exposes embeddings? no

• Tags: detection, segmentation, ocr, VLM, zero-shot

Requirements

• Packages: huggingface-hub, transformers>=4.50, torch, torchvision, accelerate, jax, flax

• CPU support

  • yes

• GPU support

  • yes

Example usage

import fiftyone as fo
import fiftyone.zoo as foz

foz.register_zoo_model_source("https://github.com/harpreetsahota204/paligemma2")

dataset = foz.load_zoo_dataset(
    "coco-2017",
    split="validation",
    dataset_name=fo.get_default_dataset_name(),
    max_samples=50,
    shuffle=True,
)

model = foz.load_zoo_model("google/paligemma2-3b-mix-448")

dataset.apply_model(model, label_field="predictions")

session = fo.launch_app(dataset)