Sockeye: A Toolkit for Neural Machine Translation

TL;DR

Sockeye is an open-source, scalable toolkit for neural machine translation supporting multiple architectures, offering competitive performance and extensive features for researchers and practitioners.

Contribution

Introduces Sockeye, a versatile NMT toolkit with support for various architectures, optimizations, and benchmarking, enhancing research and production workflows.

Findings

Sockeye achieves competitive BLEU scores on WMT benchmarks.

Transformer implementation in Sockeye outperforms other architectures.

The toolkit is open-source and facilitates research and deployment.

Abstract

We describe Sockeye (version 1.12), an open-source sequence-to-sequence toolkit for Neural Machine Translation (NMT). Sockeye is a production-ready framework for training and applying models as well as an experimental platform for researchers. Written in Python and built on MXNet, the toolkit offers scalable training and inference for the three most prominent encoder-decoder architectures: attentional recurrent neural networks, self-attentional transformers, and fully convolutional networks. Sockeye also supports a wide range of optimizers, normalization and regularization techniques, and inference improvements from current NMT literature. Users can easily run standard training recipes, explore different model settings, and incorporate new ideas. In this paper, we highlight Sockeye's features and benchmark it against other NMT toolkits on two language arcs from the 2017 Conference on Machine Translation (WMT): English-German and Latvian-English. We report competitive BLEU scores across all three architectures, including an overall best score for Sockeye's transformer implementation. To facilitate further comparison, we release all system outputs and training scripts used in our experiments. The Sockeye toolkit is free software released under the Apache 2.0 license.