florence-2-large-ft-torch

Florence-2 large-ft (0.77B) prompt-based vision foundation model fine-tuned on downstream tasks for detection, grounding, dense captioning, OCR, segmentation and captioning.

Details

• Model name: florence-2-large-ft-torch

• Model source: https://huggingface.co/florence-community/Florence-2-large-ft

• Model author: Bin Xiao, et al.

• Model license: MIT

• Model size: 1.44 GB

• Exposes embeddings? no

• Tags: detection, ocr, vlm, torch, transformer

Requirements

• Packages: torch, torchvision, transformers>=4.56.0

• CPU support

  • yes

• GPU support

  • yes

Example usage

import fiftyone as fo
import fiftyone.zoo as foz

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("florence-2-large-ft-torch")

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

session = fo.launch_app(dataset)