Note

This page was generated from the docs/examples/getting_started/torchvision_api.ipynb Jupyter notebook. You can run it locally or open it in Google Colab, see Running the Examples for more information. To prepare Colab for this exact documentation revision, run this cell before the tutorial:

setup_url = (
    "https://raw.githubusercontent.com/NVIDIA/DALI/"
    "5f8bfb8/docs/examples/colab_setup.py"
)
!curl -sSL $setup_url -o colab_setup.py
%run colab_setup.py --ref 5f8bfb8

DALI TorchVision API

Starting from DALI 2.1, an experimental TorchVision-compatible API is available in the nvidia.dali.experimental.torchvision module. It provides GPU-accelerated drop-in replacements for torchvision.transforms.v2, so existing TorchVision preprocessing code can be offloaded to DALI with minimal changes.

For the full description, the list of supported operators and instructions on implementing custom operators, see the DALI TorchVision documentation.

Drop-in replacement

Switch the import - the rest of the code stays the same:

import torch

# import torchvision.transforms.v2 as T
import nvidia.dali.experimental.torchvision as T

preprocess = T.Compose(
    [
        T.Resize(size=256),
        T.CenterCrop(size=(224, 224)),
        T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
    ]
)

image = torch.rand(3, 480, 640)  # CHW
output = preprocess(image)
print(output.shape)

torch.Size([3, 224, 224])

DALI adds two optional parameters: device ("cpu" or "gpu") for each operator and batch_size for Compose. Class-based operators must be wrapped in Compose.