PyTorch DataLoader Integration

Learn how to load TigerLabel annotation exports directly into PyTorch DataLoaders for seamless machine learning model training.

Using COCO Format with PyTorch

Export your labeled dataset in COCO format and use torchvision's built-in CocoDetection dataset class:

from torchvision.datasets import CocoDetection
from torchvision import transforms
from torch.utils.data import DataLoader

# Load COCO-format annotations from TigerLabel export
dataset = CocoDetection(
    root='path/to/images',
    annFile='path/to/annotations.json',
    transform=transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.485, 0.456, 0.406],
                           std=[0.229, 0.224, 0.225])
    ])
)

# Create DataLoader for training
dataloader = DataLoader(
    dataset,
    batch_size=16,
    shuffle=True,
    num_workers=4,
    collate_fn=lambda x: tuple(zip(*x))
)

Custom Dataset for TigerLabel Exports

For more control over data loading, create a custom PyTorch Dataset class that reads TigerLabel's JSON export format directly with your preprocessing logic.