Skip to content

🌊 Deep Lake: Multi-Modal AI Search

Why Use Deep Lake:

  • Build AI Search Applications on multi-modal data from multiple data sources.
  • Achieve high accuracy using state-of-the art search methods and query agents.
  • Scale your applications to billions of data artifacts using fast and cost-efficient search on top of object storage.

Getting Started

  1. Install Deep Lake: 

    pip install deeplake

  2. Use it in Python: 

    import deeplake
ds = deeplake.create("dataset/path")

  3. Check out our Quickstart to learn more.

For more examples, check out our RAG Quickstart and Deep Learning Quickstart.

Join our Slack Community if you need help or have suggestions for improving documentation!

What's New in Deep Lake V4

  • Multiple indexes (embedding, lexical, inverted, etc.) for fast search on object storage with minimal caching.
  • Concurrent writes with eventual consistency.
  • Faster reads/writes due to migration of low-level code to C++.