LightSeq: A High Performance Inference Library for Transformers
This work addresses the problem of efficient inference for large Transformer models, which is critical for real-world deployment in NLP applications, and represents an incremental improvement over existing methods.
The paper tackles the challenge of serving large Transformer models in industrial applications by proposing LightSeq, a high-performance inference library that achieves up to 14x speedup over TensorFlow and 1.4x over FasterTransformer on machine translation benchmarks.
Transformer, BERT and their variants have achieved great success in natural language processing. Since Transformer models are huge in size, serving these models is a challenge for real industrial applications. In this paper, we propose LightSeq, a highly efficient inference library for models in the Transformer family. LightSeq includes a series of GPU optimization techniques to to streamline the computation of neural layers and to reduce memory footprint. LightSeq can easily import models trained using PyTorch and Tensorflow. Experimental results on machine translation benchmarks show that LightSeq achieves up to 14x speedup compared with TensorFlow and 1.4x compared with FasterTransformer, a concurrent CUDA implementation. The code is available at https://github.com/bytedance/lightseq.