Timezone: »

Ouroboros: On Accelerating Training of Transformer-Based Language Models
Qian Yang · Zhouyuan Huo · Wenlin Wang · Lawrence Carin

Wed Dec 11 05:00 PM -- 07:00 PM (PST) @ East Exhibition Hall B + C #108

Language models are essential for natural language processing (NLP) tasks, such as machine translation and text summarization. Remarkable performance has been demonstrated recently across many NLP domains via a Transformer-based language model with over a billion parameters, verifying the benefits of model size. Model parallelism is required if a model is too large to fit in a single computing device. Current methods for model parallelism either suffer from backward locking in backpropagation or are not applicable to language models. We propose the first model-parallel algorithm that speeds the training of Transformer-based language models. We also prove that our proposed algorithm is guaranteed to converge to critical points for non-convex problems. Extensive experiments on Transformer and Transformer-XL language models demonstrate that the proposed algorithm obtains a much faster speedup beyond data parallelism, with comparable or better accuracy. Code to reproduce experiments is to be found at \url{https://github.com/LaraQianYang/Ouroboros}.

Author Information

Qian Yang (Duke University)
Zhouyuan Huo (University of Pittsburgh)
Wenlin Wang (Duke University)
Lawrence Carin (Duke University)

More from the Same Authors