paddle-ocr-vl-torch

PaddleOCR-VL 1.6, a 0.9B vision-language document parser that detects and reads text in one pass.

Details

• Model name: paddle-ocr-vl-torch

• Model source: https://huggingface.co/PaddlePaddle/PaddleOCR-VL-1.6

• Model author: PaddlePaddle Authors

• Model license: Apache 2.0

• Model size: 1.80 GB

• Exposes embeddings? no

• Tags: ocr, detection, torch

Requirements

• Packages: torch, transformers>=5.13

• 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("paddle-ocr-vl-torch")

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

session = fo.launch_app(dataset)