Training Models

Workflows for training models using Deep Lake datasets

How to Train Deep Learning Models Using Deep Lake

Deep Lake provides dataloaders that can be used as a drop-in replacements in existing training scripts. The benefits of Deep Lake dataloaders is their data streaming speed and compatibility with Deep Lakes query engine, which enables users to rapidly filter their data and connect it to their GPUs.

Below is a series of tutorials for training models using Deep Lake.

Training an Image Classification Model in PyTorchTraining an Object Detection and Segmentation Model in PyTorchTraining Models Using PyTorch LightningSplitting Datasets for TrainingTraining on AWS SageMakerTraining Models Using MMDetectionTraining Reproducibility Using Deep Lake and Weights & BiasesQuerying, Training and Editing Datasets with Data Lineage