DeepSpeed

DeepSpeed

DeepSpeed

Microsoft open source library

DeepSpeed is an open source deep learning optimization library for PyTorch.[1] The library is designed to reduce computing power and memory use and to train large distributed models with better parallelism on existing computer hardware.[2][3] DeepSpeed is optimized for low latency, high throughput training. It includes the Zero Redundancy Optimizer (ZeRO) for training models with 1 trillion or more parameters.[4] Features include mixed precision training, single-GPU, multi-GPU, and multi-node training as well as custom model parallelism. The DeepSpeed source code is licensed under MIT License and available on GitHub.[5]

Quick Facts Original author(s), Developer(s) ...

The team claimed to achieve up to a 6.2x throughput improvement, 2.8x faster convergence, and 4.6x less communication.[6]

See also

References

  1. [1]
    "Microsoft Updates Windows, Azure Tools with an Eye on The Future". PCMag UK. May 22, 2020.
  2. [2]
    Yegulalp, Serdar (February 10, 2020). "Microsoft speeds up PyTorch with DeepSpeed". InfoWorld.
  3. [3]
    "Microsoft unveils "fifth most powerful" supercomputer in the world". Neowin. 18 June 2023.
  4. [4]
    "Microsoft trains world's largest Transformer language model". February 10, 2020.
  5. [5]
    "microsoft/DeepSpeed". July 10, 2020 – via GitHub.
  6. [6]
    "DeepSpeed: Accelerating large-scale model inference and training via system optimizations and compression". Microsoft Research. 2021-05-24. Retrieved 2021-06-19.

Further reading

  • Rajbhandari, Samyam; Rasley, Jeff; Ruwase, Olatunji; He, Yuxiong (2019). "ZeRO: Memory Optimization Towards Training A Trillion Parameter Models". arXiv:1910.02054 [cs.LG].


Stub icon

This article about software created, produced or developed by Microsoft is a stub. You can help Wikipedia by expanding it.
Share this article:

This article uses material from the Wikipedia article DeepSpeed, and is written by contributors. Text is available under a CC BY-SA 4.0 International License; additional terms may apply. Images, videos and audio are available under their respective licenses.