Learn how to efficiently load and work with FiftyOne datasets in PyTorch using FiftyOneTorchDataset.
Train a simple PyTorch model on the MNIST dataset using \`FiftyOneTorchDataset\`, with train/validation/test splits managed in FiftyOne and a reusable training script.
Eliminate database query overhead during training by enabling vectorize mode in \`FiftyOneTorchDataset\`, preloading all required fields into memory before the training loop begins.
Create views to easily query and explore your datasets in FiftyOne.
Automatically find and remove duplicate and near-duplicate images from your FiftyOne datasets.
Check out some common workflows for finding and removing duplicate objects from your FiftyOne datasets.
Add FiftyOne to your model training and analysis loop to visualize and analyze your classifier's predictions.
Use FiftyOne to store your object detections and use the FiftyOne App to analyze them.
Render labels on the samples in your FiftyOne Dataset with a single line of code.
Use FiftyOne's powerful dataset import/export features to convert your datasets on disk between standard (or custom) formats.
Easily merge datasets on disk or in-memory using FiftyOne; e.g., to add a new set of model predictions to a dataset.
Write your own custom DatasetImporter and use it to import datasets in your custom format into FiftyOne.
Write your own custom DatasetExporter and use it to export a FiftyOne Dataset to disk in your custom format.
Write your own custom SampleParser and use it to add samples in your custom format to a FiftyOne Dataset.