Skip to content
Go to file

Latest commit

… in audio_pretraining (#1566)

Bug: `AudioPretrainingTask` is not aware of what samples have been skipped by `FileAudioDataset`, and hence would load labels of utterances that were skipped, causing `AddTargetDataset` to misalign utterances with labels. This PR tracks line line indices loaded by the `FileAudioDataset` to filter labels correspondingly in `AudioPretrainingTask`

INFO:__main__:| decoding with criterion ctc
INFO:__main__:| loading model(s) from ... 284, skipped 223 samples
INFO:__main__:| /private/home/wnhsu/wav2vec2_robust/data/joint_swbd ted_dev 284 examples
INFO:__main__:WER: 152.15605749486653
INFO:__main__:| Processed 284 sentences (72405 tokens) in 20.5s (13.88sentences/s, 3538.01 tokens/s)

INFO:__main__:| decoding with criterion ctc
INFO:__main__:| loading model(s) from ... 284, skipped 223 samples
INFO:__main__:WER: 9.904153354632587
INFO:__main__:| Processed 284 sentences (72405 tokens) in 20.7s (13.70sentences/s, 3492.20 tokens/s)

Pull Request resolved: fairinternal/fairseq-py#1566

Reviewed By: alexeib

Differential Revision: D25963317

Pulled By: wnhsu

fbshipit-source-id: c9748f5dad1ff787642ba0bc28698c4ecfbcd221

Git stats


Failed to load latest commit information.

MIT License Latest Release Build Status Documentation Status

Fairseq(-py) is a sequence modeling toolkit that allows researchers and developers to train custom models for translation, summarization, language modeling and other text generation tasks.

We provide reference implementations of various sequence modeling papers:

List of implemented papers

What's New:

Previous updates


We also provide pre-trained models for translation and language modeling with a convenient torch.hub interface:

en2de = torch.hub.load('pytorch/fairseq', 'transformer.wmt19.en-de.single_model')
en2de.translate('Hello world', beam=5)
# 'Hallo Welt'

See the PyTorch Hub tutorials for translation and RoBERTa for more examples.

Requirements and Installation

  • PyTorch version >= 1.5.0
  • Python version >= 3.6
  • For training new models, you'll also need an NVIDIA GPU and NCCL
  • To install fairseq and develop locally:
git clone
cd fairseq
pip install --editable ./

# on MacOS:
# CFLAGS="-stdlib=libc++" pip install --editable ./

# to install the latest stable release (0.10.1)
# pip install fairseq==0.10.1
  • For faster training install NVIDIA's apex library:
git clone
cd apex
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" \
  --global-option="--deprecated_fused_adam" --global-option="--xentropy" \
  --global-option="--fast_multihead_attn" ./
  • For large datasets install PyArrow: pip install pyarrow
  • If you use Docker make sure to increase the shared memory size either with --ipc=host or --shm-size as command line options to nvidia-docker run .

Getting Started

The full documentation contains instructions for getting started, training new models and extending fairseq with new model types and tasks.

Pre-trained models and examples

We provide pre-trained models and pre-processed, binarized test sets for several tasks listed below, as well as example training and evaluation commands.

We also have more detailed READMEs to reproduce results from specific papers:

Join the fairseq community


fairseq(-py) is MIT-licensed. The license applies to the pre-trained models as well.


Please cite as:

  title = {fairseq: A Fast, Extensible Toolkit for Sequence Modeling},
  author = {Myle Ott and Sergey Edunov and Alexei Baevski and Angela Fan and Sam Gross and Nathan Ng and David Grangier and Michael Auli},
  booktitle = {Proceedings of NAACL-HLT 2019: Demonstrations},
  year = {2019},
You can’t perform that action at this time.