Name: Towards AI Legal Name: Towards AI, Inc. Description: Towards AI is the world's leading artificial intelligence (AI) and technology publication. Read by thought-leaders and decision-makers around the world. Phone Number: +1-650-246-9381 Email: pub@towardsai.net
228 Park Avenue South New York, NY 10003 United States
Website: Publisher: https://towardsai.net/#publisher Diversity Policy: https://towardsai.net/about Ethics Policy: https://towardsai.net/about Masthead: https://towardsai.net/about
Name: Towards AI Legal Name: Towards AI, Inc. Description: Towards AI is the world's leading artificial intelligence (AI) and technology publication. Founders: Roberto Iriondo, , Job Title: Co-founder and Advisor Works for: Towards AI, Inc. Follow Roberto: X, LinkedIn, GitHub, Google Scholar, Towards AI Profile, Medium, ML@CMU, FreeCodeCamp, Crunchbase, Bloomberg, Roberto Iriondo, Generative AI Lab, Generative AI Lab VeloxTrend Ultrarix Capital Partners Denis Piffaretti, Job Title: Co-founder Works for: Towards AI, Inc. Louie Peters, Job Title: Co-founder Works for: Towards AI, Inc. Louis-François Bouchard, Job Title: Co-founder Works for: Towards AI, Inc. Cover:
Towards AI Cover
Logo:
Towards AI Logo
Areas Served: Worldwide Alternate Name: Towards AI, Inc. Alternate Name: Towards AI Co. Alternate Name: towards ai Alternate Name: towardsai Alternate Name: towards.ai Alternate Name: tai Alternate Name: toward ai Alternate Name: toward.ai Alternate Name: Towards AI, Inc. Alternate Name: towardsai.net Alternate Name: pub.towardsai.net
5 stars – based on 497 reviews

Frequently Used, Contextual References

TODO: Remember to copy unique IDs whenever it needs used. i.e., URL: 304b2e42315e

Resources

Our 15 AI experts built the most comprehensive, practical, 90+ lesson courses to master AI Engineering - we have pathways for any experience at Towards AI Academy. Cohorts still open - use COHORT10 for 10% off.

Publication

Grouped-Query Attention(GQA) Explained
Latest   Machine Learning

Grouped-Query Attention(GQA) Explained

Last Updated on December 30, 2023 by Editorial Team

Author(s): Florian

Originally published on Towards AI.

From Principles to Llama2 Code ExplanationGrouped-Query Attention(GQA) Explained

The standard practice for autoregressive decoding is to cache the keys and values of the previous tokens in the sequence to speed up attention computation. However, as the context window or batch size increases, the memory cost associated with the size of the key-value cache(kv cache) in the multi-head attention(MHA) model significantly increases.

Multi-Query attention(MQA) is a mechanism that uses only a single key-value head for multiple queries, which can save memory and greatly speed up decoder inference.

However, MQA may lead to a decrease in quality. In fact, we not only want fast inference, but also want the quality to be on par with MHA, so Grouped-query attention(GQA)[1] comes into play.

Grouped-query attention(GQA) is an interpolation of multi-query and multi-head attention. It achieves a quality similar to multi-head attention while maintaining a comparable speed to multi-query attention.

Since GQA is a newcomer, many famous large language models have not adopted it before. However, since its proposal, it has gained popularity among popular models such as Llama2[2] and Mistral 7B[3].

GQA can be seen as an intermediate or generalized form of MQA and MHA:

When there is only one group in GQA, it is referred to as MQA.When the number of groups in GQA is equal… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI


Take our 90+ lesson From Beginner to Advanced LLM Developer Certification: From choosing a project to deploying a working product this is the most comprehensive and practical LLM course out there!

Towards AI has published Building LLMs for Production—our 470+ page guide to mastering LLMs with practical projects and expert insights!


Discover Your Dream AI Career at Towards AI Jobs

Towards AI has built a jobs board tailored specifically to Machine Learning and Data Science Jobs and Skills. Our software searches for live AI jobs each hour, labels and categorises them and makes them easily searchable. Explore over 40,000 live jobs today with Towards AI Jobs!

Note: Content contains the views of the contributing authors and not Towards AI.