Translate is an open source project based on Facebook's machine translation systems. It uses a sequence-to-sequence model, and is based on fairseq-py, a sequence modeling toolkit for training custom models for translation, summarization, dialog, and other text generation tasks.

From research to production

Translate provides the ability to train a sequence-to-sequence model with attention, a method to export this model to Caffe2 for production using ONNX - an open format for representing deep learning models - and sample C++ code to load the exported model and run inference via beam search.

An example model is provided to demonstrate this capability. Currently, the encoder and the decoder step networks are exported separately, and beam search is implemented in C++.


Get Started

1


  • Install Translate by following the directions on GitHub.

  • 2

  • Review details on how to use examples for training, evaluating a pretrained model, exporting models to Caffe2 via ONNX, and loading and running a model from C++.

  • More Tools

    PyTorch

    PyTorch is an open source deep learning framework built to be flexible and modular for research, with the stability and support needed for production deployment. It enables fast, flexible experimentation through a tape-based autograd system designed for immediate and python-like execution.

    Fairseq

    Fairseq is a sequence modeling toolkit for training custom models for translation, summarization, and other text generation tasks.

    Join Us

    Tackle the world's most complex technology challenges

    Join Our Team

    Latest News

    Visit the AI Blog for updates on recent publications, new tools, and more.

    Visit Blog