Deep Lake: a Lakehouse for Deep Learning
It addresses the need for better data infrastructure for deep learning practitioners working with NLP, audio, computer vision, and other non-tabular datasets, though it appears incremental as an extension of existing lakehouse concepts.
The paper tackles the problem that traditional data lakes are not well-designed for deep learning applications with non-tabular data, and presents Deep Lake, an open-source lakehouse that stores complex data as tensors and streams it efficiently to deep learning frameworks while maintaining high GPU utilization.
Traditional data lakes provide critical data infrastructure for analytical workloads by enabling time travel, running SQL queries, ingesting data with ACID transactions, and visualizing petabyte-scale datasets on cloud storage. They allow organizations to break down data silos, unlock data-driven decision-making, improve operational efficiency, and reduce costs. However, as deep learning usage increases, traditional data lakes are not well-designed for applications such as natural language processing (NLP), audio processing, computer vision, and applications involving non-tabular datasets. This paper presents Deep Lake, an open-source lakehouse for deep learning applications developed at Activeloop. Deep Lake maintains the benefits of a vanilla data lake with one key difference: it stores complex data, such as images, videos, annotations, as well as tabular data, in the form of tensors and rapidly streams the data over the network to (a) Tensor Query Language, (b) in-browser visualization engine, or (c) deep learning frameworks without sacrificing GPU utilization. Datasets stored in Deep Lake can be accessed from PyTorch, TensorFlow, JAX, and integrate with numerous MLOps tools.