Introducing Optimum: The Optimization Toolkit for Transformers at Scale

📰 Hugging Face Blog

Hugging Face introduces Optimum, an optimization toolkit for scaling transformers

intermediate Published 14 Sept 2021

Action Steps

Understand the challenges of scaling transformers
Explore Optimum's features for model optimization
Quantize a model for Intel Xeon CPU using Optimum
Integrate Optimum with existing ML workflows

Who Needs to Know This

ML engineers and researchers can use Optimum to optimize transformer models for better performance and efficiency, while data scientists and software engineers can leverage it to improve model deployment and scalability

Key Insight

💡 Optimum simplifies the process of optimizing transformer models for better performance and efficiency

Key Takeaways

Hugging Face introduces Optimum, an optimization toolkit for scaling transformers

Full Article

Published Time: 2021-09-14T00:00:00.028Z

# Introducing Optimum: The Optimization Toolkit for Transformers at Scale

[![Image 1: Hugging Face's logo](https://huggingface.co/front/assets/huggingface_logo-noborder.svg)Hugging Face](https://huggingface.co/)

* [Models](https://huggingface.co/models)
* [Datasets](https://huggingface.co/datasets)
* [Spaces](https://huggingface.co/spaces)
* [Buckets new](https://huggingface.co/storage)
* [Docs](https://huggingface.co/docs)
* [Enterprise](https://huggingface.co/enterprise)
* [Pricing](https://huggingface.co/pricing)
*
*
* * *

* [Log In](https://huggingface.co/login)
* [Sign Up](https://huggingface.co/join)

[Back to Articles](https://huggingface.co/blog)

# [](https://huggingface.co/blog/hardware-partners-program#introducing-%F0%9F%A4%97-optimum-the-optimization-toolkit-for-transformers-at-scale) Introducing 🤗 Optimum: The Optimization Toolkit for Transformers at Scale

Published September 14, 2021

[Update on GitHub](https://github.com/huggingface/blog/blob/main/hardware-partners-program.md)

[- [x] Upvote 2](https://huggingface.co/login?next=%2Fblog%2Fhardware-partners-program)
* [![Image 2](https://cdn-avatars.huggingface.co/v1/production/uploads/6425839a6d0f0f5f1dc6e344/01SIMX66NEDFScQa2DBR4.png)](https://huggingface.co/cindyangelira "cindyangelira")
* [![Image 3](https://cdn-avatars.huggingface.co/v1/production/uploads/6640bbd0220cfa8cbfdce080/wiAHUu5ewawyipNs0YFBR.png)](https://huggingface.co/John6666 "John6666")

[![Image 4: Morgan Funtowicz's avatar](https://cdn-avatars.huggingface.co/v1/production/uploads/1583858935715-5e67c47c100906368940747e.jpeg)](https://huggingface.co/mfuntowicz)

[Morgan Funtowicz mfuntowicz Follow](https://huggingface.co/mfuntowicz)

[![Image 5: Ella Charlaix's avatar](https://cdn-avatars.huggingface.co/v1/production/uploads/1615915889033-6050eb5aeb94f56898c08e57.jpeg)](https://huggingface.co/echarlaix)

[Ella Charlaix echarlaix Follow](https://huggingface.co/echarlaix)

[![Image 6: Michael Benayoun's avatar](https://cdn-avatars.huggingface.co/v1/production/uploads/1615890856777-6047a3315da6ba4b1dfb9e18.png)](https://huggingface.co/michaelbenayoun)

[Michael Benayoun michaelbenayoun Follow](https://huggingface.co/michaelbenayoun)

[![Image 7: Jeff Boudier's avatar](https://cdn-avatars.huggingface.co/v1/production/uploads/1605114051380-noauth.jpeg)](https://huggingface.co/jeffboudier)

[Jeff Boudier jeffboudier Follow](https://huggingface.co/jeffboudier)

* [Why 🤗 Optimum?](https://huggingface.co/blog/hardware-partners-program#why-%F0%9F%A4%97-optimum "Why 🤗 Optimum?")
* [🤯 Scaling Transformers is hard](https://huggingface.co/blog/hardware-partners-program#%F0%9F%A4%AF-scaling-transformers-is-hard "🤯 Scaling Transformers is hard")

* [🏭 Optimum puts Transformers to work](https://huggingface.co/blog/hardware-partners-program#%F0%9F%8F%AD-optimum-puts-transformers-to-work "🏭 Optimum puts Transformers to work")

* [🤗 Optimum in practice: how to quantize a model for Intel Xeon CPU](https://huggingface.co/blog/hardware-partners-program#%F0%9F%A4%97-optimum-in-practice-how-to-quantize-a-model-for-intel-xeon-cpu "🤗 Optimum in practice: how to quantize a model for Intel Xeon CPU")
* [🤔 Why quantization is important but tricky to get right](https://huggingface.co/blog/hardware-partners-program#%F0%9F%A4%94-why-quantization-is-important-but-tricky-to-get-right "🤔 Why quantization is important but tricky to get right")

* [💡 How Intel is solving quantization and more with Neural Compressor](https://huggingface.co/blog/hardware-partners-program#%F0%9F%92%A1-how-intel-is-solving-quantization-and-more-with-neural-compressor "💡 How Intel is solving quantization and more with Neural Compressor")

* [🔥 How to easily quantize Transformers for Intel Xeon CPUs with Optimum](https://huggingface.co/blog/hardware-partners-program#%F0%9F%94%A5-how-to-easily-quantize-transformers-for-intel-xeon-cpus-

Read full article → ← Back to Reads

Introducing Optimum: The Optimization Toolkit for Transformers at Scale

Key Takeaways

Full Article

Related Videos