Counting 1,868 Big Data & Machine Learning Frameworks, Toolsets, and Examples...
Suggestion? Feedback? Tweet @stkim1

Last Commit
Oct. 18, 2017
Aug. 29, 2017


FAIR Sequence-to-Sequence Toolkit (PyTorch)

This is a PyTorch version of fairseq, a sequence-to-sequence learning toolkit from Facebook AI Research. The original authors of this reimplementation are (in no particular order) Sergey Edunov, Myle Ott, and Sam Gross. The toolkit implements the fully convolutional model described in Convolutional Sequence to Sequence Learning and features multi-GPU training on a single machine as well as fast beam search generation on both CPU and GPU. We provide pre-trained models for English to French and English to German translation.



If you use the code in your paper, then please cite it as:

  author    = {Gehring, Jonas, and Auli, Michael and Grangier, David and Yarats, Denis and Dauphin, Yann N},
  title     = "{Convolutional Sequence to Sequence Learning}",
  booktitle = {Proc. of ICML},
  year      = 2017,

Requirements and Installation

  • A computer running macOS or Linux
  • For training new models, you'll also need a NVIDIA GPU and NCCL
  • Python version 3.6
  • A PyTorch installation

Currently fairseq-py requires installing PyTorch from source. Please follow the instructions here:

After PyTorch is installed, you can install fairseq-py with:

pip install -r requirements.txt
python build
python develop

Quick Start

The following command-line tools are available:

  • python Data pre-processing: build vocabularies and binarize training data
  • python Train a new model on one or multiple GPUs
  • python Translate pre-processed data with a trained model
  • python -i: Translate raw text with a trained model
  • python BLEU scoring of generated translations against reference translations

Evaluating Pre-trained Models

First, download a pre-trained model along with its vocabularies:

$ curl | tar xvjf -

This model uses a Byte Pair Encoding (BPE) vocabulary, so we'll have to apply the encoding to the source text before it can be translated. This can be done with the script using the wmt14.en-fr.fconv-cuda/bpecodes file. @@ is used as a continuation marker and the original text can be easily recovered with e.g. sed s/@@ //g or by passing the --remove-bpe flag to Prior to BPE, input text needs to be tokenized using tokenizer.perl from mosesdecoder.

Let's use python -i to generate translations. Here, we use a beam size of 5:

$ MODEL_DIR=wmt14.en-fr.fconv-py
$ python -i \
 --path $MODEL_DIR/ $MODEL_DIR \
 --beam 5
| [en] dictionary: 44206 types
| [fr] dictionary: 44463 types
| model fconv_wmt_en_fr
| loaded checkpoint /private/home/edunov/wmt14.en-fr.fconv-py/ (epoch 37)
> Why is it rare to discover new marine [email protected]@ mal species ?
S       Why is it rare to discover new marine [email protected]@ mal species ?
O       Why is it rare to discover new marine [email protected]@ mal species ?
H       -0.08662842959165573    Pourquoi est-il rare de découvrir de nouvelles espèces de mammifères marins ?
A       0 1 3 3 5 6 6 10 8 8 8 11 12

This generation script produces four types of outputs: a line prefixed with S shows the supplied source sentence after applying the vocabulary; O is a copy of the original source sentence; H is the hypothesis along with an average log-likelihood; and A is the attention maxima for each word in the hypothesis, including the end-of-sentence marker which is omitted from the text.

Check below for a full list of pre-trained models available.

Training a New Model

Data Pre-processing

The fairseq-py source distribution contains an example pre-processing script for the IWSLT 2014 German-English corpus. Pre-process and binarize the data as follows:

$ cd data/
$ bash
$ cd ..
$ TEXT=data/
$ python --source-lang de --target-lang en \
  --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
  --destdir data-bin/

This will write binarized data that can be used for model training to data-bin/


Use python to train a new model. Here a few example settings that work well for the IWSLT 2014 dataset:

$ mkdir -p checkpoints/fconv
$ CUDA_VISIBLE_DEVICES=0 python data-bin/ \
  --lr 0.25 --clip-norm 0.1 --dropout 0.2 --max-tokens 4000 \
  --arch fconv_iwslt_de_en --save-dir checkpoints/fconv

By default, python will use all available GPUs on your machine. Use the CUDA_VISIBLE_DEVICES environment variable to select specific GPUs and/or to change the number of GPU devices that will be used.

Also note that the batch size is specified in terms of the maximum number of tokens per batch (--max-tokens). You may need to use a smaller value depending on the available GPU memory on your system.


Once your model is trained, you can generate translations using python (for binarized data) or python -i (for raw text):

$ python data-bin/ \
  --path checkpoints/fconv/ \
  --batch-size 128 --beam 5
  | [de] dictionary: 35475 types
  | [en] dictionary: 24739 types
  | data-bin/ test 6750 examples
  | model fconv
  | loaded checkpoint trainings/fconv/
  S-721   danke .
  T-721   thank you .

To generate translations with only a CPU, use the --cpu flag. BPE continuation markers can be removed with the --remove-bpe flag.

Pre-trained Models

We provide the following pre-trained fully convolutional sequence-to-sequence models:

In addition, we provide pre-processed and binarized test sets for the models above:

Generation with the binarized test sets can be run in batch mode as follows, e.g. for English-French on a GTX-1080ti:

$ curl | tar xvjf - -C data-bin
$ curl | tar xvjf - -C data-bin
$ python data-bin/wmt14.en-fr.newstest2014  \
  --path data-bin/wmt14.en-fr.fconv-py/ \
  --beam 5 --batch-size 128 --remove-bpe | tee /tmp/gen.out
| Translated 3003 sentences (95451 tokens) in 94.5s (1009.60 tokens/s)
| Generate test with beam=5: BLEU4 = 40.70, 67.7/46.8/34.2/25.4 (BP=1.000, ratio=1.000, syslen=81190, reflen=81194)

# Scoring with
$ grep ^H /tmp/gen.out | cut -f3- > /tmp/gen.out.sys
$ grep ^T /tmp/gen.out | cut -f2- > /tmp/gen.out.ref
$ python --sys /tmp/gen.out.sys --ref /tmp/gen.out.ref
BLEU4 = 40.70, 67.7/46.8/34.2/25.4 (BP=1.000, ratio=1.000, syslen=81190, reflen=81194)

Join the fairseq community


fairseq-py is BSD-licensed. The license applies to the pre-trained models as well. We also provide an additional patent grant.