DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale

Samyam Rajbhandari, Conglong Li, Zhewei Yao, Minjia Zhang, Reza Yazdani Aminabadi, Ammar Ahmad Awan, Jeff Rasley, Yuxiong He
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:18332-18346, 2022.

Abstract

As the training of giant dense models hits the boundary on the availability and capability of the hardware resources today, Mixture-of-Experts (MoE) models have become one of the most promising model architectures due to their significant training cost reduction compared to quality-equivalent dense models. Their training cost saving is demonstrated from encoder-decoder models (prior works) to a 5x saving for auto-aggressive language models (this work). However, due to the much larger model size and unique architecture, how to provide fast MoE model inference remains challenging and unsolved, limiting their practical usage. To tackle this, we present DeepSpeed-MoE, an end-to-end MoE training and inference solution, including novel MoE architecture designs and model compression techniques that reduce MoE model size by up to 3.7x, and a highly optimized inference system that provides 7.3x better latency and cost compared to existing MoE inference solutions. DeepSpeed-MoE offers an unprecedented scale and efficiency to serve massive MoE models with up to 4.5x faster and 9x cheaper inference compared to quality-equivalent dense models. We hope our innovations and systems help open a promising path to new directions in the large model landscape, a shift from dense to sparse MoE models, where training and deploying higher-quality models with fewer resources becomes more widely possible.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-rajbhandari22a, title = {{D}eep{S}peed-{M}o{E}: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation {AI} Scale}, author = {Rajbhandari, Samyam and Li, Conglong and Yao, Zhewei and Zhang, Minjia and Aminabadi, Reza Yazdani and Awan, Ammar Ahmad and Rasley, Jeff and He, Yuxiong}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {18332--18346}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/rajbhandari22a/rajbhandari22a.pdf}, url = {https://proceedings.mlr.press/v162/rajbhandari22a.html}, abstract = {As the training of giant dense models hits the boundary on the availability and capability of the hardware resources today, Mixture-of-Experts (MoE) models have become one of the most promising model architectures due to their significant training cost reduction compared to quality-equivalent dense models. Their training cost saving is demonstrated from encoder-decoder models (prior works) to a 5x saving for auto-aggressive language models (this work). However, due to the much larger model size and unique architecture, how to provide fast MoE model inference remains challenging and unsolved, limiting their practical usage. To tackle this, we present DeepSpeed-MoE, an end-to-end MoE training and inference solution, including novel MoE architecture designs and model compression techniques that reduce MoE model size by up to 3.7x, and a highly optimized inference system that provides 7.3x better latency and cost compared to existing MoE inference solutions. DeepSpeed-MoE offers an unprecedented scale and efficiency to serve massive MoE models with up to 4.5x faster and 9x cheaper inference compared to quality-equivalent dense models. We hope our innovations and systems help open a promising path to new directions in the large model landscape, a shift from dense to sparse MoE models, where training and deploying higher-quality models with fewer resources becomes more widely possible.} }
Endnote
%0 Conference Paper %T DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale %A Samyam Rajbhandari %A Conglong Li %A Zhewei Yao %A Minjia Zhang %A Reza Yazdani Aminabadi %A Ammar Ahmad Awan %A Jeff Rasley %A Yuxiong He %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-rajbhandari22a %I PMLR %P 18332--18346 %U https://proceedings.mlr.press/v162/rajbhandari22a.html %V 162 %X As the training of giant dense models hits the boundary on the availability and capability of the hardware resources today, Mixture-of-Experts (MoE) models have become one of the most promising model architectures due to their significant training cost reduction compared to quality-equivalent dense models. Their training cost saving is demonstrated from encoder-decoder models (prior works) to a 5x saving for auto-aggressive language models (this work). However, due to the much larger model size and unique architecture, how to provide fast MoE model inference remains challenging and unsolved, limiting their practical usage. To tackle this, we present DeepSpeed-MoE, an end-to-end MoE training and inference solution, including novel MoE architecture designs and model compression techniques that reduce MoE model size by up to 3.7x, and a highly optimized inference system that provides 7.3x better latency and cost compared to existing MoE inference solutions. DeepSpeed-MoE offers an unprecedented scale and efficiency to serve massive MoE models with up to 4.5x faster and 9x cheaper inference compared to quality-equivalent dense models. We hope our innovations and systems help open a promising path to new directions in the large model landscape, a shift from dense to sparse MoE models, where training and deploying higher-quality models with fewer resources becomes more widely possible.
APA
Rajbhandari, S., Li, C., Yao, Z., Zhang, M., Aminabadi, R.Y., Awan, A.A., Rasley, J. & He, Y.. (2022). DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:18332-18346 Available from https://proceedings.mlr.press/v162/rajbhandari22a.html.

Related Material