update

fairseq/.github/ISSUE_TEMPLATE.md

@@ -0,0 +1,3 @@
+## 👉 [Please follow one of these issue templates](https://github.com/pytorch/fairseq/issues/new/choose) 👈
+
+Note: to keep the backlog clean and actionable, issues may be immediately closed if they do not follow one of the above issue templates.

fairseq/.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,43 @@
+---
+name: 🐛 Bug Report
+about: Submit a bug report to help us improve
+labels: 'bug, needs triage'
+---
+
+## 🐛 Bug
+
+<!-- A clear and concise description of what the bug is. -->
+
+### To Reproduce
+
+Steps to reproduce the behavior (**always include the command you ran**):
+
+1. Run cmd '....'
+2. See error
+
+<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
+
+
+#### Code sample
+<!-- Ideally attach a minimal code sample to reproduce the decried issue.
+Minimal means having the shortest code but still preserving the bug. -->
+
+### Expected behavior
+
+<!-- A clear and concise description of what you expected to happen. -->
+
+### Environment
+
+ - fairseq Version (e.g., 1.0 or main):
+ - PyTorch Version (e.g., 1.0)
+ - OS (e.g., Linux):
+ - How you installed fairseq (`pip`, source):
+ - Build command you used (if compiling from source):
+ - Python version:
+ - CUDA/cuDNN version:
+ - GPU models and configuration:
+ - Any other relevant information:
+
+### Additional context
+
+<!-- Add any other context about the problem here. -->

fairseq/.github/ISSUE_TEMPLATE/documentation.md

@@ -0,0 +1,15 @@
+---
+name: 📚 Documentation/Typos
+about: Report an issue related to documentation or a typo
+labels: 'documentation, needs triage'
+---
+
+## 📚 Documentation
+
+For typos and doc fixes, please go ahead and:
+
+1. Create an issue.
+2. Fix the typo.
+3. Submit a PR.
+
+Thanks!

fairseq/.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,24 @@
+---
+name: 🚀 Feature Request
+about: Submit a proposal/request for a new feature
+labels: 'enhancement, help wanted, needs triage'
+---
+
+## 🚀 Feature Request
+<!-- A clear and concise description of the feature proposal -->
+
+### Motivation
+
+<!-- Please outline the motivation for the proposal. Is your feature request related to a problem? e.g., I'm always frustrated when [...]. If this is related to another GitHub issue, please link here too -->
+
+### Pitch
+
+<!-- A clear and concise description of what you want to happen. -->
+
+### Alternatives
+
+<!-- A clear and concise description of any alternative solutions or features you've considered, if any. -->
+
+### Additional context
+
+<!-- Add any other context or screenshots about the feature request here. -->

fairseq/.github/ISSUE_TEMPLATE/how-to-question.md

@@ -0,0 +1,33 @@
+---
+name: ❓ Questions/Help
+about: If you have questions, please first search existing issues and docs
+labels: 'question, needs triage'
+---
+
+## ❓ Questions and Help
+
+### Before asking:
+1. search the issues.
+2. search the docs.
+
+<!-- If you still can't find what you need: -->
+
+#### What is your question?
+
+#### Code
+
+<!-- Please paste a code snippet if your question requires it! -->
+
+#### What have you tried?
+
+#### What's your environment?
+
+ - fairseq Version (e.g., 1.0 or main):
+ - PyTorch Version (e.g., 1.0)
+ - OS (e.g., Linux):
+ - How you installed fairseq (`pip`, source):
+ - Build command you used (if compiling from source):
+ - Python version:
+ - CUDA/cuDNN version:
+ - GPU models and configuration:
+ - Any other relevant information:

fairseq/.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,16 @@
+# Before submitting
+
+- [ ] Was this discussed/approved via a Github issue? (no need for typos, doc improvements)
+- [ ] Did you read the [contributor guideline](https://github.com/pytorch/fairseq/blob/main/CONTRIBUTING.md)?
+- [ ] Did you make sure to update the docs?
+- [ ] Did you write any new necessary tests?
+
+## What does this PR do?
+Fixes # (issue).
+
+## PR review
+Anyone in the community is free to review the PR once the tests have passed.
+If we didn't discuss your PR in Github issues there's a high chance it will not be merged.
+
+## Did you have fun?
+Make sure you had fun coding 🙃

fairseq/.github/stale.yml

@@ -0,0 +1,30 @@
+# Configuration for probot-stale - https://github.com/probot/stale
+# Mostly copied from github.com/facebook/react/blob/master/.github/stale.yml
+# Number of days of inactivity before an issue becomes stale
+daysUntilStale: 90
+# Number of days of inactivity before a stale issue is closed
+daysUntilClose: 7
+# Issues with these labels will never be considered stale
+exemptLabels:
+  - bug
+# Label to use when marking an issue as stale
+staleLabel: stale
+issues:
+  # Comment to post when marking an issue as stale.
+  markComment: >
+    This issue has been automatically marked as stale.
+    **If this issue is still affecting you, please leave any comment** (for example, "bump"), and we'll keep it open.
+    We are sorry that we haven't been able to prioritize it yet. If you have any new additional information, please include it with your comment!
+  # Comment to post when closing a stale issue.
+  closeComment: >
+    Closing this issue after a prolonged period of inactivity. If this issue is still present in the latest release, please create a new issue with up-to-date information. Thank you!
+pulls:
+  # Comment to post when marking a pull request as stale.
+  markComment: >
+    This pull request has been automatically marked as stale.
+    **If this pull request is still relevant, please leave any comment** (for example, "bump"), and we'll keep it open.
+    We are sorry that we haven't been able to prioritize reviewing it yet. Your contribution is very much appreciated.
+  # Comment to post when closing a stale pull request.
+  closeComment: >
+    Closing this pull request after a prolonged period of inactivity. If this issue is still present in the latest release, please ask for this pull request to be reopened. Thank you!
+

fairseq/.github/workflows/build.yml

@@ -0,0 +1,55 @@
+name: build
+
+on:
+  # Trigger the workflow on push to main or any pull request
+  push:
+    branches:
+      - main
+  pull_request:
+
+jobs:
+  build:
+
+    strategy:
+      max-parallel: 4
+      matrix:
+        platform: [ubuntu-latest, macos-latest]
+        python-version: [3.6, 3.7]
+
+    runs-on: ${{ matrix.platform }}
+
+    steps:
+    - uses: actions/checkout@v2
+
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v2
+      with:
+        python-version: ${{ matrix.python-version }}
+
+    - name: Conditionally install pytorch
+      if: matrix.platform == 'windows-latest'
+      run: pip3 install torch -f https://download.pytorch.org/whl/torch_stable.html
+
+    - name: Install locally
+      run: |
+        python -m pip install --upgrade pip
+        git submodule update --init --recursive
+        python setup.py build_ext --inplace
+        python -m pip install --editable .
+
+    - name: Install optional test requirements
+      run: |
+        python -m pip install iopath transformers pyarrow
+        python -m pip install git+https://github.com/facebookresearch/fairscale.git@main
+
+    - name: Lint with flake8
+      run: |
+        pip install flake8
+        # stop the build if there are Python syntax errors or undefined names
+        flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics --extend-exclude fairseq/model_parallel/megatron
+        # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
+        flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics --extend-exclude fairseq/model_parallel/megatron
+
+    - name: Run tests
+      run: |
+          python setup.py test

fairseq/.github/workflows/build_wheels.yml

@@ -0,0 +1,41 @@
+name: build_wheels
+
+on:
+  push:
+    branches:
+      - v[0-9]+.[0-9]+.[x0-9]+
+    tags:
+      - v*
+
+jobs:
+  build_wheels:
+    name: Build wheels on ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, macos-latest]
+
+    steps:
+      - uses: actions/checkout@v2
+
+      - name: Install Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: '3.7'
+
+      - name: Install cibuildwheel
+        run: |
+          python -m pip install cibuildwheel
+
+      - name: Build wheels for CPython
+        run: |
+          python -m cibuildwheel --output-dir dist
+        env:
+          CIBW_BUILD: "cp36-*64 cp37-*64 cp38-*64"
+          CIBW_MANYLINUX_X86_64_IMAGE: manylinux1
+          CIBW_BEFORE_BUILD: git submodule update --init --recursive && pip install .
+
+      - uses: actions/upload-artifact@v2
+        with:
+          name: wheels
+          path: ./dist/*.whl

fairseq/.gitmodules

@@ -0,0 +1,4 @@
+[submodule "fairseq/model_parallel/megatron"]
+    path = fairseq/model_parallel/megatron
+    url = https://github.com/ngoyal2707/Megatron-LM
+    branch = fairseq

fairseq/CODE_OF_CONDUCT.md

@@ -0,0 +1,77 @@
+# Code of Conduct
+
+## Our Pledge
+
+In the interest of fostering an open and welcoming environment, we as
+contributors and maintainers pledge to make participation in our project and
+our community a harassment-free experience for everyone, regardless of age, body
+size, disability, ethnicity, sex characteristics, gender identity and expression,
+level of experience, education, socio-economic status, nationality, personal
+appearance, race, religion, or sexual identity and orientation.
+
+## Our Standards
+
+Examples of behavior that contributes to creating a positive environment
+include:
+
+* Using welcoming and inclusive language
+* Being respectful of differing viewpoints and experiences
+* Gracefully accepting constructive criticism
+* Focusing on what is best for the community
+* Showing empathy towards other community members
+
+Examples of unacceptable behavior by participants include:
+
+* The use of sexualized language or imagery and unwelcome sexual attention or
+  advances
+* Trolling, insulting/derogatory comments, and personal or political attacks
+* Public or private harassment
+* Publishing others' private information, such as a physical or electronic
+  address, without explicit permission
+* Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Our Responsibilities
+
+Project maintainers are responsible for clarifying the standards of acceptable
+behavior and are expected to take appropriate and fair corrective action in
+response to any instances of unacceptable behavior.
+
+Project maintainers have the right and responsibility to remove, edit, or
+reject comments, commits, code, wiki edits, issues, and other contributions
+that are not aligned to this Code of Conduct, or to ban temporarily or
+permanently any contributor for other behaviors that they deem inappropriate,
+threatening, offensive, or harmful.
+
+## Scope
+
+This Code of Conduct applies within all project spaces, and it also applies when
+an individual is representing the project or its community in public spaces.
+Examples of representing a project or community include using an official
+project e-mail address, posting via an official social media account, or acting
+as an appointed representative at an online or offline event. Representation of
+a project may be further defined and clarified by project maintainers.
+
+## Enforcement
+
+Instances of abusive, harassing, or otherwise unacceptable behavior may be
+reported by contacting the project team at <[email protected]>. All
+complaints will be reviewed and investigated and will result in a response that
+is deemed necessary and appropriate to the circumstances. The project team is
+obligated to maintain confidentiality with regard to the reporter of an incident.
+Further details of specific enforcement policies may be posted separately.
+
+Project maintainers who do not follow or enforce the Code of Conduct in good
+faith may face temporary or permanent repercussions as determined by other
+members of the project's leadership.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 1.4,
+available at https://www.contributor-covenant.org/version/1/4/code-of-conduct.html
+
+[homepage]: https://www.contributor-covenant.org
+
+For answers to common questions about this code of conduct, see
+https://www.contributor-covenant.org/faq
+

fairseq/CONTRIBUTING.md

@@ -0,0 +1,28 @@
+# Contributing to Facebook AI Research Sequence-to-Sequence Toolkit (fairseq)
+We want to make contributing to this project as easy and transparent as
+possible.
+
+## Pull Requests
+We actively welcome your pull requests.
+
+1. Fork the repo and create your branch from `main`.
+2. If you've added code that should be tested, add tests.
+3. If you've changed APIs, update the documentation.
+4. Ensure the test suite passes.
+5. Make sure your code lints.
+6. If you haven't already, complete the Contributor License Agreement ("CLA").
+
+## Contributor License Agreement ("CLA")
+In order to accept your pull request, we need you to submit a CLA. You only need
+to do this once to work on any of Facebook's open source projects.
+
+Complete your CLA here: <https://code.facebook.com/cla>
+
+## Issues
+We use GitHub issues to track public bugs. Please ensure your description is
+clear and has sufficient instructions to be able to reproduce the issue.
+
+## License
+By contributing to Facebook AI Research Sequence-to-Sequence Toolkit (fairseq),
+you agree that your contributions will be licensed under the LICENSE file in
+the root directory of this source tree.

fairseq/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) Facebook, Inc. and its affiliates.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

fairseq/README.md

@@ -0,0 +1,229 @@
+<p align="center">
+  <img src="docs/fairseq_logo.png" width="150">
+  <br />
+  <br />
+  <a href="https://github.com/pytorch/fairseq/blob/main/LICENSE"><img alt="MIT License" src="https://img.shields.io/badge/license-MIT-blue.svg" /></a>
+  <a href="https://github.com/pytorch/fairseq/releases"><img alt="Latest Release" src="https://img.shields.io/github/release/pytorch/fairseq.svg" /></a>
+  <a href="https://github.com/pytorch/fairseq/actions?query=workflow:build"><img alt="Build Status" src="https://github.com/pytorch/fairseq/workflows/build/badge.svg" /></a>
+  <a href="https://fairseq.readthedocs.io/en/latest/?badge=latest"><img alt="Documentation Status" src="https://readthedocs.org/projects/fairseq/badge/?version=latest" /></a>
+</p>
+
+--------------------------------------------------------------------------------
+
+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:
+
+<details><summary>List of implemented papers</summary><p>
+
+* **Convolutional Neural Networks (CNN)**
+  + [Language Modeling with Gated Convolutional Networks (Dauphin et al., 2017)](examples/language_model/conv_lm/README.md)
+  + [Convolutional Sequence to Sequence Learning (Gehring et al., 2017)](examples/conv_seq2seq/README.md)
+  + [Classical Structured Prediction Losses for Sequence to Sequence Learning (Edunov et al., 2018)](https://github.com/pytorch/fairseq/tree/classic_seqlevel)
+  + [Hierarchical Neural Story Generation (Fan et al., 2018)](examples/stories/README.md)
+  + [wav2vec: Unsupervised Pre-training for Speech Recognition (Schneider et al., 2019)](examples/wav2vec/README.md)
+* **LightConv and DynamicConv models**
+  + [Pay Less Attention with Lightweight and Dynamic Convolutions (Wu et al., 2019)](examples/pay_less_attention_paper/README.md)
+* **Long Short-Term Memory (LSTM) networks**
+  + Effective Approaches to Attention-based Neural Machine Translation (Luong et al., 2015)
+* **Transformer (self-attention) networks**
+  + Attention Is All You Need (Vaswani et al., 2017)
+  + [Scaling Neural Machine Translation (Ott et al., 2018)](examples/scaling_nmt/README.md)
+  + [Understanding Back-Translation at Scale (Edunov et al., 2018)](examples/backtranslation/README.md)
+  + [Adaptive Input Representations for Neural Language Modeling (Baevski and Auli, 2018)](examples/language_model/README.adaptive_inputs.md)
+  + [Lexically constrained decoding with dynamic beam allocation (Post & Vilar, 2018)](examples/constrained_decoding/README.md)
+  + [Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context (Dai et al., 2019)](examples/truncated_bptt/README.md)
+  + [Adaptive Attention Span in Transformers (Sukhbaatar et al., 2019)](examples/adaptive_span/README.md)
+  + [Mixture Models for Diverse Machine Translation: Tricks of the Trade (Shen et al., 2019)](examples/translation_moe/README.md)
+  + [RoBERTa: A Robustly Optimized BERT Pretraining Approach (Liu et al., 2019)](examples/roberta/README.md)
+  + [Facebook FAIR's WMT19 News Translation Task Submission (Ng et al., 2019)](examples/wmt19/README.md)
+  + [Jointly Learning to Align and Translate with Transformer Models (Garg et al., 2019)](examples/joint_alignment_translation/README.md )
+  + [Multilingual Denoising Pre-training for Neural Machine Translation (Liu et at., 2020)](examples/mbart/README.md)
+  + [Neural Machine Translation with Byte-Level Subwords (Wang et al., 2020)](examples/byte_level_bpe/README.md)
+  + [Unsupervised Quality Estimation for Neural Machine Translation (Fomicheva et al., 2020)](examples/unsupervised_quality_estimation/README.md)
+  + [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations (Baevski et al., 2020)](examples/wav2vec/README.md)
+  + [Generating Medical Reports from Patient-Doctor Conversations Using Sequence-to-Sequence Models (Enarvi et al., 2020)](examples/pointer_generator/README.md)
+  + [Linformer: Self-Attention with Linear Complexity (Wang et al., 2020)](examples/linformer/README.md)
+  + [Cross-lingual Retrieval for Iterative Self-Supervised Training (Tran et al., 2020)](examples/criss/README.md)
+  + [Deep Transformers with Latent Depth (Li et al., 2020)](examples/latent_depth/README.md)
+  + [Unsupervised Cross-lingual Representation Learning for Speech Recognition (Conneau et al., 2020)](https://arxiv.org/abs/2006.13979)
+  + [Robust wav2vec 2.0: Analyzing Domain Shift in Self-Supervised Pre-Training (Hsu, et al., 2021)](https://arxiv.org/abs/2104.01027)
+  + [Unsupervised Speech Recognition (Baevski, et al., 2021)](https://arxiv.org/abs/2105.11084)
+* **Non-autoregressive Transformers**
+  + Non-Autoregressive Neural Machine Translation (Gu et al., 2017)
+  + Deterministic Non-Autoregressive Neural Sequence Modeling by Iterative Refinement (Lee et al. 2018)
+  + Insertion Transformer: Flexible Sequence Generation via Insertion Operations (Stern et al. 2019)
+  + Mask-Predict: Parallel Decoding of Conditional Masked Language Models (Ghazvininejad et al., 2019)
+  + [Levenshtein Transformer (Gu et al., 2019)](examples/nonautoregressive_translation/README.md)
+* **Finetuning**
+  + [Better Fine-Tuning by Reducing Representational Collapse (Aghajanyan et al. 2020)](examples/rxf/README.md)
+
+</p></details>
+
+### What's New:
+
+* September 2021 [`master` branch renamed to `main`](https://github.com/github/renaming).
+* July 2021 [Released DrNMT code](examples/discriminative_reranking_nmt/README.md)
+* July 2021 [Released Robust wav2vec 2.0 model](examples/wav2vec/README.md)
+* June 2021 [Released XLMR-XL and XLMR-XXL models](examples/xlmr/README.md)
+* May 2021 [Released Unsupervised Speech Recognition code](examples/wav2vec/unsupervised/README.md)
+* March 2021 [Added full parameter and optimizer state sharding + CPU offloading](examples/fully_sharded_data_parallel/README.md)
+* February 2021 [Added LASER training code](examples/laser/README.md)
+* December 2020: [Added Adaptive Attention Span code](examples/adaptive_span/README.md)
+* December 2020: [GottBERT model and code released](examples/gottbert/README.md)
+* November 2020: Adopted the [Hydra](https://github.com/facebookresearch/hydra) configuration framework
+  * [see documentation explaining how to use it for new and existing projects](docs/hydra_integration.md)
+* November 2020: [fairseq 0.10.0 released](https://github.com/pytorch/fairseq/releases/tag/v0.10.0)
+* October 2020: [Added R3F/R4F (Better Fine-Tuning) code](examples/rxf/README.md)
+* October 2020: [Deep Transformer with Latent Depth code released](examples/latent_depth/README.md)
+* October 2020: [Added CRISS models and code](examples/criss/README.md)
+
+<details><summary>Previous updates</summary><p>
+
+* September 2020: [Added Linformer code](examples/linformer/README.md)
+* September 2020: [Added pointer-generator networks](examples/pointer_generator/README.md)
+* August 2020: [Added lexically constrained decoding](examples/constrained_decoding/README.md)
+* August 2020: [wav2vec2 models and code released](examples/wav2vec/README.md)
+* July 2020: [Unsupervised Quality Estimation code released](examples/unsupervised_quality_estimation/README.md)
+* May 2020: [Follow fairseq on Twitter](https://twitter.com/fairseq)
+* April 2020: [Monotonic Multihead Attention code released](examples/simultaneous_translation/README.md)
+* April 2020: [Quant-Noise code released](examples/quant_noise/README.md)
+* April 2020: [Initial model parallel support and 11B parameters unidirectional LM released](examples/megatron_11b/README.md)
+* March 2020: [Byte-level BPE code released](examples/byte_level_bpe/README.md)
+* February 2020: [mBART model and code released](examples/mbart/README.md)
+* February 2020: [Added tutorial for back-translation](https://github.com/pytorch/fairseq/tree/main/examples/backtranslation#training-your-own-model-wmt18-english-german)
+* December 2019: [fairseq 0.9.0 released](https://github.com/pytorch/fairseq/releases/tag/v0.9.0)
+* November 2019: [VizSeq released (a visual analysis toolkit for evaluating fairseq models)](https://facebookresearch.github.io/vizseq/docs/getting_started/fairseq_example)
+* November 2019: [CamemBERT model and code released](examples/camembert/README.md)
+* November 2019: [BART model and code released](examples/bart/README.md)
+* November 2019: [XLM-R models and code released](examples/xlmr/README.md)
+* September 2019: [Nonautoregressive translation code released](examples/nonautoregressive_translation/README.md)
+* August 2019: [WMT'19 models released](examples/wmt19/README.md)
+* July 2019: fairseq relicensed under MIT license
+* July 2019: [RoBERTa models and code released](examples/roberta/README.md)
+* June 2019: [wav2vec models and code released](examples/wav2vec/README.md)
+
+</p></details>
+
+### Features:
+
+* multi-GPU training on one machine or across multiple machines (data and model parallel)
+* fast generation on both CPU and GPU with multiple search algorithms implemented:
+  + beam search
+  + Diverse Beam Search ([Vijayakumar et al., 2016](https://arxiv.org/abs/1610.02424))
+  + sampling (unconstrained, top-k and top-p/nucleus)
+  + [lexically constrained decoding](examples/constrained_decoding/README.md) (Post & Vilar, 2018)
+* [gradient accumulation](https://fairseq.readthedocs.io/en/latest/getting_started.html#large-mini-batch-training-with-delayed-updates) enables training with large mini-batches even on a single GPU
+* [mixed precision training](https://fairseq.readthedocs.io/en/latest/getting_started.html#training-with-half-precision-floating-point-fp16) (trains faster with less GPU memory on [NVIDIA tensor cores](https://developer.nvidia.com/tensor-cores))
+* [extensible](https://fairseq.readthedocs.io/en/latest/overview.html): easily register new models, criterions, tasks, optimizers and learning rate schedulers
+* [flexible configuration](docs/hydra_integration.md) based on [Hydra](https://github.com/facebookresearch/hydra) allowing a combination of code, command-line and file based configuration
+* [full parameter and optimizer state sharding](examples/fully_sharded_data_parallel/README.md)
+* [offloading parameters to CPU](examples/fully_sharded_data_parallel/README.md)
+
+We also provide [pre-trained models for translation and language modeling](#pre-trained-models-and-examples)
+with a convenient `torch.hub` interface:
+
+``` python
+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](https://pytorch.org/hub/pytorch_fairseq_translation/)
+and [RoBERTa](https://pytorch.org/hub/pytorch_fairseq_roberta/) for more examples.
+
+# Requirements and Installation
+
+* [PyTorch](http://pytorch.org/) version >= 1.5.0
+* Python version >= 3.6
+* For training new models, you'll also need an NVIDIA GPU and [NCCL](https://github.com/NVIDIA/nccl)
+* **To install fairseq** and develop locally:
+
+``` bash
+git clone https://github.com/pytorch/fairseq
+cd fairseq
+pip install --editable ./
+
+# on MacOS:
+# CFLAGS="-stdlib=libc++" pip install --editable ./
+
+# to install the latest stable release (0.10.x)
+# pip install fairseq
+```
+
+* **For faster training** install NVIDIA's [apex](https://github.com/NVIDIA/apex) library:
+
+``` bash
+git clone https://github.com/NVIDIA/apex
+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](https://arrow.apache.org/docs/python/install.html#using-pip): `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](https://fairseq.readthedocs.io/) 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.
+
+* [Translation](examples/translation/README.md): convolutional and transformer models are available
+* [Language Modeling](examples/language_model/README.md): convolutional and transformer models are available
+
+We also have more detailed READMEs to reproduce results from specific papers:
+
+* [Cross-lingual Retrieval for Iterative Self-Supervised Training (Tran et al., 2020)](examples/criss/README.md)
+* [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations (Baevski et al., 2020)](examples/wav2vec/README.md)
+* [Unsupervised Quality Estimation for Neural Machine Translation (Fomicheva et al., 2020)](examples/unsupervised_quality_estimation/README.md)
+* [Training with Quantization Noise for Extreme Model Compression ({Fan*, Stock*} et al., 2020)](examples/quant_noise/README.md)
+* [Neural Machine Translation with Byte-Level Subwords (Wang et al., 2020)](examples/byte_level_bpe/README.md)
+* [Multilingual Denoising Pre-training for Neural Machine Translation (Liu et at., 2020)](examples/mbart/README.md)
+* [Reducing Transformer Depth on Demand with Structured Dropout (Fan et al., 2019)](examples/layerdrop/README.md)
+* [Jointly Learning to Align and Translate with Transformer Models (Garg et al., 2019)](examples/joint_alignment_translation/README.md)
+* [Levenshtein Transformer (Gu et al., 2019)](examples/nonautoregressive_translation/README.md)
+* [Facebook FAIR's WMT19 News Translation Task Submission (Ng et al., 2019)](examples/wmt19/README.md)
+* [RoBERTa: A Robustly Optimized BERT Pretraining Approach (Liu et al., 2019)](examples/roberta/README.md)
+* [wav2vec: Unsupervised Pre-training for Speech Recognition (Schneider et al., 2019)](examples/wav2vec/README.md)
+* [Mixture Models for Diverse Machine Translation: Tricks of the Trade (Shen et al., 2019)](examples/translation_moe/README.md)
+* [Pay Less Attention with Lightweight and Dynamic Convolutions (Wu et al., 2019)](examples/pay_less_attention_paper/README.md)
+* [Understanding Back-Translation at Scale (Edunov et al., 2018)](examples/backtranslation/README.md)
+* [Classical Structured Prediction Losses for Sequence to Sequence Learning (Edunov et al., 2018)](https://github.com/pytorch/fairseq/tree/classic_seqlevel)
+* [Hierarchical Neural Story Generation (Fan et al., 2018)](examples/stories/README.md)
+* [Scaling Neural Machine Translation (Ott et al., 2018)](examples/scaling_nmt/README.md)
+* [Convolutional Sequence to Sequence Learning (Gehring et al., 2017)](examples/conv_seq2seq/README.md)
+* [Language Modeling with Gated Convolutional Networks (Dauphin et al., 2017)](examples/language_model/README.conv.md)
+
+# Join the fairseq community
+
+* Twitter: https://twitter.com/fairseq
+* Facebook page: https://www.facebook.com/groups/fairseq.users
+* Google group: https://groups.google.com/forum/#!forum/fairseq-users
+
+# License
+
+fairseq(-py) is MIT-licensed.
+The license applies to the pre-trained models as well.
+
+# Citation
+
+Please cite as:
+
+``` bibtex
+@inproceedings{ott2019fairseq,
+  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},
+}
+```

fairseq/examples/__init__.py

@@ -0,0 +1,9 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+try:
+    from fairseq.version import __version__  # noqa
+except ImportError:
+    pass

fairseq/examples/adaptive_span/README.md

@@ -0,0 +1,90 @@
+# Adaptive Span
+
+Adaptive Span is a novel self-attention mechanism that can learn its optimal
+attention span. This allows us to extend significantly the maximum context size
+used in Transformer, while maintaining control over their memory footprint
+and computational time. It uses the Truncated BPTT technique for training,
+as in [transformerXL](https://github.com/pytorch/fairseq/blob/main/examples/truncated_bptt/README.md).
+
+Adaptive Span was introduced by paper:
+[Adaptive Attention Span in Transformers](https://arxiv.org/abs/1905.07799),
+which achieved state-of-the-art language modeling results at the time of publication.
+
+We manage to reproduce their result in fairseq and keep most of the
+[original implementation](https://github.com/facebookresearch/adaptive-span) untouched.
+You can refer to the their sweep file as well if any combination of hyperparameter is not clear.
+
+##### 0. Setup
+
+First you need to process the Enwik8 dataset, we use the pre-tokenized dataset
+from [adaptive span paper](https://github.com/facebookresearch/adaptive-span/blob/master/get_data.sh).
+You can download the dataset, and then run:
+```bash
+fairseq-preprocess --only-source --trainpref ~/data/enwik8/train.txt \
+    --validpref ~/data/enwik8/valid.txt --testpref ~/data/enwik8/test.txt \
+    --destdir ~/data/enwik8/data-bin/ --joined-dictionary --workers 20
+```
+
+##### 1. Train a Adaptive Span model on Enwik8
+
+We will train a 12-layer Adaptive Span model following the [hyperparameters
+used in the original
+paper](https://github.com/facebookresearch/adaptive-span/blob/master/experiments/enwik8.sh).
+
+The following command assumes 4 GPUs, so that the total batch size is 64
+sequences (4 x 16). Training should take 2-3 days on 4 V100 GPUs:
+```bash
+CUDA_VISIBLE_DEVICES=0,1,2,3 fairseq-train \
+    --user-dir examples/adaptive_span \
+    --data  ~/data/enwik8/data-bin/ \
+    --fp16 --fp16-no-flatten-grads --max-update 600000 \
+    --task truncated_bptt_lm --tokens-per-sample 512 --arch adaptive_span \
+    --n-layer 12 --d-model 512 --n-head 8 --d-inner 2048 --dropout 0.3 \
+    --attn-span 8192 --optimizer adagrad_with_grad_clip --adagrad-clip 0.03 \
+    --validate-interval-updates 1000 \
+    --lr-scheduler fixed --warmup-updates 32000 --batch-size-valid 32 \
+    --lr 0.07 --criterion adaptive_span_loss --batch-size 16 --update-freq 1 \
+    --seed 2 --log-format json --log-interval 25 --aux-loss-scaler 5e-07
+```
+This should land around 1.05 on validation, 1.03 on test. You can lower the
+--aux-loss-scaler for better performance (longer span). It gives ~0.03 bpc
+improvement to the transformerXL baseline here.
+If training on a single GPU, set `--update-freq=4` to accumulate 4x gradients
+and simulate training on 4 GPUs.
+You can also reproduce the transformerXL result on enwik8 using this code base.
+It should land around 1.06 on test,matching the [original paper](https://github.com/kimiyoung/transformer-xl/blob/master/pytorch/run_enwik8_base.sh).
+You can try by
+```bash
+CUDA_VISIBLE_DEVICES=0,1,2,3 fairseq-train \
+    --user-dir examples/truncated_bptt \
+    ~/data/enwik8/data-bin/ \
+    --task truncated_bptt_lm  --fp16 --max-update 400000 \
+    --tokens-per-sample 512 --arch transformer_xl --n-layer 12 \
+    --d-model 512 --n-head 8 --d-head 64 --d-inner 2048 --dropout 0.1 \
+    --dropatt 0.0 --mem-len 512 --optimizer adam --clip-norm 0.25 \
+    --lr-scheduler cosine --warmup-updates 0 \
+    --lr 0.0 --lr 0.00025 --batch-size 15 \
+    --update-freq 1 --seed 2 --log-format json --log-interval 25 \
+    --fp16
+```
+
+##### 2. Evaluate
+For Adaptive Span:
+```bash
+fairseq-eval-lm ~/data/enwik8/data-bin/ --path model/checkpoint_best.pt \
+ --user-dir examples/adaptive_span \
+ --task truncated_bptt_lm --batch-size 8 --tokens-per-sample 512 --gen-subset test
+```
+For Transformer-XL evaluation:
+```bash
+fairseq-eval-lm ~/data/enwik8/data-bin/ --path model/checkpoint_best.pt \
+    --user-dir examples/truncated_bptt/ --task truncated_bptt_lm --batch-size 8 \
+    --tokens-per-sample 80 \
+    --model-overrides '{"mem_len":2100,"clamp_len":820,"same_length":True}' \
+    --gen-subset valid
+```
+
+*Note:* During training the model saw 512 tokens of context
+(``--tokens-per-sample=512``), with batch size 8. These settings match the evaluation
+settings from [the original
+paper](https://github.com/facebookresearch/adaptive-span/blob/master/experiments/enwik8.sh).

fairseq/examples/adaptive_span/__init__.py

@@ -0,0 +1,19 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import importlib
+import os
+
+# automatically import any Python files in the current directory
+cur_dir = os.path.dirname(__file__)
+for file in os.listdir(cur_dir):
+    path = os.path.join(cur_dir, file)
+    if (
+        not file.startswith("_")
+        and not file.startswith(".")
+        and (file.endswith(".py") or os.path.isdir(path))
+    ):
+        mod_name = file[: file.find(".py")] if file.endswith(".py") else file
+        module = importlib.import_module(__name__ + "." + mod_name)

fairseq/examples/adaptive_span/adagrad_with_grad_clip.py

@@ -0,0 +1,128 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+from torch.optim import Adagrad
+
+from fairseq.optim import LegacyFairseqOptimizer, register_optimizer
+
+
+@register_optimizer("adagrad_with_grad_clip")
+class FairseqAdagradWithGradClip(LegacyFairseqOptimizer):
+    def __init__(self, args, params):
+        super().__init__(args)
+        self._optimizer = AdagradWithGradClip(params, **self.optimizer_config)
+
+    @staticmethod
+    def add_args(parser):
+        """Add optimizer-specific arguments to the parser."""
+        # fmt: off
+        parser.add_argument('--weight-decay', '--wd', default=0.0, type=float, metavar='WD',
+                            help='weight decay')
+        parser.add_argument('--adagrad-clip', default=0.0, type=float, metavar='D',
+                            help='internal grad clip')
+        # fmt: on
+
+    @property
+    def optimizer_config(self):
+        """
+        Return a kwarg dictionary that will be used to override optimizer
+        args stored in checkpoints. This allows us to load a checkpoint and
+        resume training using a different set of optimizer args, e.g., with a
+        different learning rate.
+        """
+        return {
+            "lr": self.args.lr[0],
+            "weight_decay": self.args.weight_decay,
+            "grad_clip": self.args.adagrad_clip,
+        }
+
+    @property
+    def supports_flat_params(self):
+        return False
+
+
+def _clip_grad(clr, grad, group_grad_clip):
+    if group_grad_clip > 0:
+        norm = grad.norm(2).item()
+        if norm > group_grad_clip:
+            clr *= group_grad_clip / (norm + 1e-10)
+    return clr
+
+
+class AdagradWithGradClip(Adagrad):
+    """Adagrad algorithm with custom gradient clipping"""
+
+    def __init__(
+        self,
+        params,
+        lr=1e-2,
+        lr_decay=0,
+        weight_decay=0,
+        initial_accumulator_value=0,
+        grad_clip=0,
+    ):
+        Adagrad.__init__(
+            self,
+            params,
+            lr=lr,
+            lr_decay=lr_decay,
+            weight_decay=weight_decay,
+            initial_accumulator_value=initial_accumulator_value,
+        )
+        self.defaults["grad_clip"] = grad_clip
+        self.param_groups[0].setdefault("grad_clip", grad_clip)
+
+    def step(self, closure=None):
+        loss = None
+        if closure is not None:
+            loss = closure()
+
+        for group in self.param_groups:
+            for p in group["params"]:
+                if p.grad is None:
+                    continue
+
+                grad = p.grad.data
+                state = self.state[p]
+
+                state["step"] += 1
+
+                if group["weight_decay"] != 0:
+                    if p.grad.data.is_sparse:
+                        raise RuntimeError(
+                            "weight_decay option is "
+                            "not compatible with sparse "
+                            "gradients"
+                        )
+                    grad = grad.add(group["weight_decay"], p.data)
+
+                clr = group["lr"] / (1 + (state["step"] - 1) * group["lr_decay"])
+
+                # clip
+                clr = _clip_grad(clr=clr, grad=grad, group_grad_clip=group["grad_clip"])
+
+                if grad.is_sparse:
+                    # the update is non-linear so indices must be unique
+                    grad = grad.coalesce()
+                    grad_indices = grad._indices()
+                    grad_values = grad._values()
+                    size = grad.size()
+
+                    def make_sparse(values):
+                        constructor = grad.new
+                        if grad_indices.dim() == 0 or values.dim() == 0:
+                            return constructor().resize_as_(grad)
+                        return constructor(grad_indices, values, size)
+
+                    state["sum"].add_(make_sparse(grad_values.pow(2)))
+                    std = state["sum"]._sparse_mask(grad)
+                    std_values = std._values().sqrt_().add_(1e-10)
+                    p.data.add_(-clr, make_sparse(grad_values / std_values))
+                else:
+                    state["sum"].addcmul_(1, grad, grad)
+                    std = state["sum"].sqrt().add_(1e-10)
+                    p.data.addcdiv_(-clr, grad, std)
+
+        return loss

fairseq/examples/adaptive_span/adaptive_span_attention.py

@@ -0,0 +1,160 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+import math
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class AdaptiveMask(nn.Module):
+    """Soft masking function for adaptive size.
+    It masks out the last K values of an input. The masking value
+    goes from 1 to 0 gradually, so K can be learned with
+    back-propagation.
+    Args:
+        max_size: maximum size (i.e. input dimension)
+        ramp_size: size of the ramp going from 0 to 1
+        init_val: initial size proportion not to be masked out
+        shape: learn multiple sizes independent of each other
+    """
+
+    def __init__(self, max_size, ramp_size, init_val=0, shape=(1,)):
+        nn.Module.__init__(self)
+        self._max_size = max_size
+        self._ramp_size = ramp_size
+        self.current_val = nn.Parameter(torch.zeros(*shape) + init_val)
+        mask_template = torch.linspace(1 - max_size, 0, steps=max_size)
+        self.register_buffer("mask_template", mask_template)
+
+    def forward(self, x):
+        mask = self.mask_template.float() + self.current_val.float() * self._max_size
+        mask = mask / self._ramp_size + 1
+        mask = mask.clamp(0, 1)
+        if x.size(-1) < self._max_size:
+            # the input could have been trimmed beforehand to save computation
+            mask = mask.narrow(-1, self._max_size - x.size(-1), x.size(-1))
+        x = (x * mask).type_as(x)
+        return x
+
+    def get_current_max_size(self, include_ramp=True):
+        current_size = math.ceil(self.current_val.max().item() * self._max_size)
+        if include_ramp:
+            current_size += self._ramp_size
+        current_size = max(0, min(self._max_size, current_size))
+        return current_size
+
+    def get_current_avg_size(self, include_ramp=True):
+        current_size = math.ceil(
+            self.current_val.float().mean().item() * self._max_size
+        )
+        if include_ramp:
+            current_size += self._ramp_size
+        current_size = max(0, min(self._max_size, current_size))
+        return current_size
+
+    def clamp_param(self):
+        """this need to be called after each update"""
+        self.current_val.data.clamp_(0, 1)
+
+
+class AdaptiveSpan(nn.Module):
+    """Adaptive attention span for Transformerself.
+    This module learns an attention span length from data for each
+    self-attention head.
+    Args:
+        attn_span: maximum attention span
+        adapt_span_loss: loss coefficient for the span length
+        adapt_span_ramp: length of the masking ramp
+        adapt_span_init: initial size ratio
+        adapt_span_cache: adapt cache size to reduce memory usage
+    """
+
+    def __init__(
+        self,
+        attn_span,
+        adapt_span_ramp,
+        adapt_span_init,
+        n_head,
+        adapt_span_layer,
+        **kargs
+    ):
+        nn.Module.__init__(self)
+        self._max_span = attn_span
+        self._n_head = n_head
+        self._adapt_span_layer = adapt_span_layer
+        if self._adapt_span_layer:
+            self._mask = AdaptiveMask(
+                max_size=self._max_span,
+                ramp_size=adapt_span_ramp,
+                init_val=adapt_span_init,
+            )
+        else:
+            self._mask = AdaptiveMask(
+                max_size=self._max_span,
+                ramp_size=adapt_span_ramp,
+                init_val=adapt_span_init,
+                shape=(n_head, 1, 1),
+            )
+
+    def forward(self, attn, normalize=True):
+        """mask attention with the right span"""
+        # batch and head dimensions are merged together, so separate them first
+        self.clamp_param()
+        if self._adapt_span_layer:
+            attn = self._mask(attn)
+        else:
+            B = attn.size(0)  # batch size
+            M = attn.size(1)  # block size
+            attn = attn.reshape(B // self._n_head, self._n_head, M, -1)
+            attn = self._mask(attn)
+            attn = attn.view(B, M, -1)
+        return attn
+
+    def get_trim_len(self):
+        """how much of memory can be trimmed to reduce computation"""
+        L = self._max_span
+        trim_len = min(L - 1, L - self._mask.get_current_max_size())
+        # too fine granularity might be bad for the memory management
+        trim_len = math.floor(trim_len / 64) * 64
+        return trim_len
+
+    def trim_memory(self, query, key, value, key_pe):
+        """trim out unnecessary memory beforehand to reduce computation"""
+        trim_len = self.get_trim_len()
+        cache_size = key.size(1) - query.size(1)
+        trim_len_cache = trim_len - (self._max_span - cache_size)
+        if trim_len_cache > 0:
+            key = key[:, trim_len_cache:, :]
+            value = value[:, trim_len_cache:, :]
+        elif trim_len_cache < 0:
+            # cache is too short! this happens when validation resumes
+            # after a lot of updates.
+            key = F.pad(key, [0, 0, -trim_len_cache, 0])
+            value = F.pad(value, [0, 0, -trim_len_cache, 0])
+        if trim_len > 0:
+            if key_pe is not None:
+                key_pe = key_pe[:, :, trim_len:]
+        return key, value, key_pe
+
+    def get_cache_size(self):
+        """determine how long the cache should be"""
+        trim_len = self.get_trim_len()
+        # give a buffer of 64 steps since a span might increase
+        # in future updates
+        return min(self._max_span, self._max_span - trim_len + 64)
+
+    def get_loss(self):
+        """a loss term for regularizing the span length"""
+        return self._max_span * self._mask.current_val.float().mean()
+
+    def get_current_max_span(self):
+        return self._mask.get_current_max_size()
+
+    def get_current_avg_span(self):
+        return self._mask.get_current_avg_size()
+
+    def clamp_param(self):
+        self._mask.clamp_param()

fairseq/examples/adaptive_span/adaptive_span_loss.py

@@ -0,0 +1,106 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+from dataclasses import dataclass
+
+import torch.nn.functional as F
+from fairseq import metrics, utils
+from fairseq.criterions import register_criterion
+from fairseq.criterions.cross_entropy import CrossEntropyCriterion
+from fairseq.dataclass import FairseqDataclass
+from omegaconf import II
+
+
+@dataclass
+class AdaptiveSpanCriterionConfig(FairseqDataclass):
+    sentence_avg: bool = II("optimization.sentence_avg")
+
+
+@register_criterion("adaptive_span_loss", dataclass=AdaptiveSpanCriterionConfig)
+class AdaptiveSpanCriterion(CrossEntropyCriterion):
+    def __init__(self, task, sentence_avg):
+        super().__init__(task, sentence_avg)
+
+    def forward(self, model, sample, reduce=True):
+        """Compute the loss for the given sample.
+
+        Returns a tuple with three elements:
+        1) the loss here is summed, different from the adaptive span code
+        2) the sample size, which is used as the denominator for the gradient
+        3) logging outputs to display while training
+        """
+        net_output = model(**sample["net_input"])
+        loss, aux_loss, avg_span, max_span = self.compute_loss(
+            model, net_output, sample, reduce=reduce
+        )
+        sample_size = (
+            sample["target"].size(0) if self.sentence_avg else sample["ntokens"]
+        )
+        loss /= sample_size
+        total_loss = loss + aux_loss
+        sample_size = 1
+
+        logging_output = {
+            "loss": loss.data,
+            "ntokens": sample["ntokens"],
+            "nsentences": sample["target"].size(0),
+            "sample_size": sample_size,
+            "total_loss": total_loss.data,
+            "avg_span": avg_span * sample_size,
+            "max_span": max_span * sample_size,
+        }
+        return total_loss, sample_size, logging_output
+
+    def compute_loss(self, model, net_output, sample, reduce=True):
+        loss, _ = super().compute_loss(model, net_output, sample, reduce)
+        aux_loss = model.get_aux_loss()
+        avg_span = model.get_current_avg_span()
+        max_span = model.get_current_max_span()
+        return loss, aux_loss, avg_span, max_span
+
+    @staticmethod
+    def reduce_metrics(logging_outputs) -> None:
+        """Aggregate logging outputs from data parallel training."""
+        loss_sum = sum(log.get("loss", 0) for log in logging_outputs)
+        ntokens = sum(log.get("ntokens", 0) for log in logging_outputs)
+        sample_size = sum(log.get("sample_size", 0) for log in logging_outputs)
+        total_loss_sum = sum(log.get("total_loss", 0) for log in logging_outputs)
+        avg_span_sum = sum(log.get("avg_span", 0) for log in logging_outputs)
+        max_span_sum = sum(log.get("max_span", 0) for log in logging_outputs)
+
+        # we divide by log(2) to convert the loss from base e to base 2
+        metrics.log_scalar(
+            "loss", loss_sum / sample_size / math.log(2), sample_size, round=3
+        )
+        metrics.log_scalar("avg_span", avg_span_sum / sample_size, sample_size, round=3)
+        metrics.log_scalar("max_span", max_span_sum / sample_size, sample_size, round=3)
+        # total loss contains the L1 norm on adaptive-span
+        metrics.log_scalar(
+            "total_loss",
+            total_loss_sum / sample_size / math.log(2),
+            sample_size,
+            round=3,
+        )
+        if sample_size != ntokens:
+            metrics.log_scalar(
+                "nll_loss", loss_sum / ntokens / math.log(2), ntokens, round=3
+            )
+            metrics.log_derived(
+                "ppl", lambda meters: utils.get_perplexity(meters["nll_loss"].avg)
+            )
+        else:
+            metrics.log_derived(
+                "ppl", lambda meters: utils.get_perplexity(meters["loss"].avg)
+            )
+
+    @staticmethod
+    def logging_outputs_can_be_summed() -> bool:
+        """
+        Whether the logging outputs returned by `forward` can be summed
+        across workers prior to calling `reduce_metrics`. Setting this
+        to True will improves distributed training speed.
+        """
+        return True

fairseq/examples/adaptive_span/adaptive_span_model.py

@@ -0,0 +1,263 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from fairseq.modules.layer_norm import LayerNorm
+
+from .adaptive_span_attention import AdaptiveSpan
+
+# Size notations:
+# B = batch_size, H = d_model, M = block_size, L = attn_span
+
+
+def _skew(X, pad_value):
+    """shift every row 1 step to right"""
+    # X = B x M x L
+    B, M, L = X.size()
+    X = F.pad(X, (0, M + 1), value=pad_value)  # B x M x (L+M+1)
+    X = X.view(B, -1)  # B x ML+MM+M
+    X = X[:, :-M]  # B x ML+MM
+    X = X.view(B, M, M + L)  # B x M x L+M
+    return X
+
+
+def _unskew(X):
+    """reverse _skew operation"""
+    # X = B x M x L+M
+    B, M, L = X.size()
+    L -= M
+    X = X.view(B, -1)  # B x ML+MM
+    X = F.pad(X, (0, M))  # B x ML+MM+M
+    X = X.view(B, M, M + L + 1)  # B x M x L+M+1
+    X = X[:, :, :L]  # B x M x L
+    return X
+
+
+class SeqAttention(nn.Module):
+    """Sequential self-attention layer.
+    Each token will attend to its previous fixed number of steps.
+    Note that attention doesn't include the current step itself.
+    """
+
+    def __init__(self, d_model, n_head, attn_span, dropout, adapt_span_layer, **kargs):
+        nn.Module.__init__(self)
+        self.dropout = nn.Dropout(dropout)
+        self.d_model = d_model  # size of a single head
+        self.attn_span = attn_span
+        self.adaptive_span = AdaptiveSpan(
+            attn_span=attn_span,
+            n_head=n_head,
+            adapt_span_layer=adapt_span_layer,
+            **kargs
+        )
+
+    def forward(self, query, key, value, key_pe):
+        # query size = B x M x H
+        # key, value sizes = B x (M+L) x H
+
+        key, value, key_pe = self.adaptive_span.trim_memory(query, key, value, key_pe)
+
+        # compute attention from context
+        # B x M (dest) x (M+L) (src)
+        attn_cont = torch.matmul(query, key.transpose(-1, -2))
+        attn_cont = _unskew(attn_cont)  # B x M x L
+
+        # compute the effect of position embedding
+        attn_pos = torch.matmul(query, key_pe)  # B x M x L_pos
+        attn = attn_cont + attn_pos
+
+        attn = attn / math.sqrt(self.d_model)  # B x M X L_pos
+
+        attn = F.softmax(attn.float(), dim=-1).type_as(attn)
+
+        # trim attention lengths according to the learned span
+        attn = self.adaptive_span(attn)
+
+        attn = self.dropout(attn)  # B x M X L_pos
+
+        attn_cont = _skew(attn, 0)  # B x M X (L+M)
+        out = torch.matmul(attn_cont, value)  # B x M x H
+        return out
+
+    def get_cache_size(self):
+        return self.adaptive_span.get_cache_size()
+
+
+class MultiHeadSeqAttention(nn.Module):
+    def __init__(self, d_model, n_head, **kargs):
+        nn.Module.__init__(self)
+        assert d_model % n_head == 0
+        self.n_head = n_head
+        self.head_dim = d_model // n_head
+        self.attn = SeqAttention(d_model=self.head_dim, n_head=n_head, **kargs)
+        self.proj_query = nn.Linear(d_model, d_model, bias=False)
+        nn.init.xavier_normal_(self.proj_query.weight)
+        self.proj_out = nn.Linear(d_model, d_model, bias=False)
+        nn.init.xavier_normal_(self.proj_out.weight)
+        self.proj_val = nn.Linear(d_model, d_model, bias=False)
+        nn.init.xavier_normal_(self.proj_val.weight)
+        self.proj_key = nn.Linear(d_model, d_model, bias=False)
+        nn.init.xavier_normal_(self.proj_key.weight)
+
+    def head_reshape(self, x):
+        K = self.n_head
+        D = self.head_dim
+        x = x.view(x.size()[:-1] + (K, D))  # B x (M+L) x K x D
+        x = x.transpose(1, 2).contiguous()  # B x K x (M+L) x D
+        x = x.view(-1, x.size(-2), x.size(-1))  # B_K x (M+L) x D
+        return x
+
+    def forward(self, query, key, value, key_pe):
+        B = query.size(0)
+        K = self.n_head
+        D = self.head_dim
+        M = query.size(1)
+
+        query = self.proj_query(query)
+        query = self.head_reshape(query)
+        value = self.proj_val(value)
+        value = self.head_reshape(value)
+        key = self.proj_key(key)
+        key = self.head_reshape(key)
+
+        out = self.attn(query, key, value, key_pe)  # B_K x M x D
+        out = out.view(B, K, M, D)  # B x K x M x D
+        out = out.transpose(1, 2).contiguous()  # B x M x K x D
+        out = out.view(B, M, -1)  # B x M x K_D
+        out = self.proj_out(out)
+        return out
+
+
+class FeedForwardLayer(nn.Module):
+    def __init__(self, d_model, d_inner, dropout, **kargs):
+        nn.Module.__init__(self)
+        self.fc1 = nn.Linear(d_model, d_inner)
+        self.fc2 = nn.Linear(d_inner, d_model)
+        nn.init.xavier_uniform_(self.fc1.weight)
+        nn.init.xavier_uniform_(self.fc2.weight)
+        self.dropout = nn.Dropout(dropout)
+
+    def forward(self, h):
+        h1 = F.relu(self.fc1(h))
+        h1 = self.dropout(h1)
+        h2 = self.fc2(h1)
+        return h2
+
+
+class TransformerSeqLayer(nn.Module):
+    def __init__(self, d_model, **kargs):
+        nn.Module.__init__(self)
+        self.attn = MultiHeadSeqAttention(d_model=d_model, **kargs)
+        self.norm1 = LayerNorm(d_model)
+        self.ff = FeedForwardLayer(d_model=d_model, **kargs)
+        self.norm2 = LayerNorm(d_model)
+
+    def forward(self, h, h_cache, key_pe):
+        # h = B x M x H
+        # h_cache = B x L x H
+        h_all = torch.cat([h_cache, h], dim=1)  # B x (M+L) x H
+        attn_out = self.attn(h, h_all, h_all, key_pe)
+        h = self.norm1(h + attn_out)  # B x M x H
+        if self.ff is not None:
+            ff_out = self.ff(h)
+            out = self.norm2(h + ff_out)  # B x M x H
+        else:
+            out = h
+        return out
+
+    def get_cache_size(self):
+        return self.attn.attn.get_cache_size()
+
+
+class TransformerSeq(nn.Module):
+    def __init__(
+        self,
+        vocab_size,
+        d_model,
+        n_head,
+        n_layer,
+        attn_span,
+        emb_dropout,
+        aux_loss_scaler,
+        adapt_span_layer,
+        **kargs
+    ):
+        nn.Module.__init__(self)
+        # token embeddings
+        self.in_emb = nn.Embedding(vocab_size, d_model)
+        nn.init.normal_(self.in_emb.weight, mean=0, std=d_model ** -0.5)
+        self.out_emb = nn.Linear(d_model, vocab_size)
+        self.aux_loss_scaler = aux_loss_scaler
+        if emb_dropout > 0:
+            self.emb_dropout = nn.Dropout(emb_dropout)
+        else:
+            self.emb_dropout = None
+        # position embeddings
+        self.key_pe = nn.Parameter(torch.randn(1, d_model // n_head, attn_span))
+
+        self.layers = nn.ModuleList()
+        self.layers.extend(
+            TransformerSeqLayer(
+                d_model=d_model,
+                n_head=n_head,
+                attn_span=attn_span,
+                adapt_span_layer=adapt_span_layer,
+                **kargs
+            )
+            for _ in range(n_layer)
+        )
+
+    def forward(self, x, h_cache, target=None):
+        # x size = B x M
+        block_size = x.size(1)
+        h = self.in_emb(x)  # B x M x H
+        if self.emb_dropout is not None:
+            h = self.emb_dropout(h)
+
+        h_cache_next = []
+        for l, layer in enumerate(self.layers):
+            cache_size = layer.attn.attn.get_cache_size()
+            if cache_size > block_size:
+                h_cache_next_l = torch.cat(
+                    [h_cache[l][:, -cache_size + block_size :, :], h], dim=1
+                ).detach()
+            else:
+                h_cache_next_l = h[:, -cache_size:, :].detach()
+            h_cache_next.append(h_cache_next_l)
+            h = layer(h, h_cache[l], self.key_pe)  # B x M x H
+
+        if self.emb_dropout is not None:
+            h = self.emb_dropout(h)
+
+        out = F.log_softmax(self.out_emb(h).float(), dim=-1).type_as(h)
+        dummy_loss = None
+
+        return out, h_cache_next, dummy_loss
+
+    def get_aux_loss(self):
+        loss = 0.0
+        for layer in self.layers:
+            loss += layer.attn.attn.adaptive_span.get_loss()
+        return self.aux_loss_scaler * loss
+
+    def get_current_max_span(self):
+        max_span = 0.0
+        for layer in self.layers:
+            max_span = max(
+                max_span, layer.attn.attn.adaptive_span.get_current_max_span()
+            )
+        return max_span
+
+    def get_current_avg_span(self):
+        avg_span = 0.0
+        for layer in self.layers:
+            avg_span += layer.attn.attn.adaptive_span.get_current_avg_span()
+        return avg_span / len(self.layers)

fairseq/examples/adaptive_span/adaptive_span_model_wrapper.py

@@ -0,0 +1,145 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+from dataclasses import dataclass
+from typing import Dict, List, Optional
+
+import torch
+from fairseq.dataclass import FairseqDataclass
+from fairseq.models import (
+    FairseqIncrementalDecoder,
+    FairseqLanguageModel,
+    register_model,
+)
+from .adaptive_span_model import TransformerSeq as AdaptiveSpanTransformerModel
+
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class AdaptiveSpanSmallConfig(FairseqDataclass):
+    # defaults come from https://github.com/facebookresearch/adaptive-span/blob/master/experiments/enwik8_small.sh
+    vocab_size: int = 50
+    d_model: int = 256
+    n_head: int = 4
+    d_inner: int = 1024
+    n_layer: int = 8
+    attn_span: int = 1024
+    dropout: float = 0.0
+    emb_dropout: float = 0.0
+    adapt_span_ramp: int = 32
+    adapt_span_init: float = 0.0
+    aux_loss_scaler: float = 0.000002
+    adapt_span_layer: bool = False
+
+
+@register_model("adaptive_span", dataclass=AdaptiveSpanSmallConfig)
+class AdaptiveSpanTransformer(FairseqLanguageModel):
+    @classmethod
+    def build_model(cls, cfg: AdaptiveSpanSmallConfig, task):
+        return cls(AdaptiveSpanDecoder(cfg, task))
+
+    def get_aux_loss(self):
+        return self.decoder.get_aux_loss()
+
+    def get_current_max_span(self):
+        return self.decoder.get_current_max_span()
+
+    def get_current_avg_span(self):
+        return self.decoder.get_current_avg_span()
+
+
+class AdaptiveSpanDecoder(FairseqIncrementalDecoder):
+    def __init__(self, cfg, task):
+
+        super().__init__(task.target_dictionary)
+
+        self.config = cfg
+        config = AdaptiveSpanSmallConfig(
+            vocab_size=len(task.target_dictionary),
+            d_model=cfg.d_model,
+            n_head=cfg.n_head,
+            d_inner=cfg.d_inner,
+            n_layer=cfg.n_layer,
+            attn_span=cfg.attn_span,
+            dropout=cfg.dropout,
+            emb_dropout=cfg.emb_dropout,
+            adapt_span_ramp=cfg.adapt_span_ramp,
+            adapt_span_init=cfg.adapt_span_init,
+            aux_loss_scaler=cfg.aux_loss_scaler,
+            adapt_span_layer=cfg.adapt_span_layer,
+        )
+        logger.info(config)
+        self.model = AdaptiveSpanTransformerModel(**config.__dict__)
+
+        self._mems = None
+
+    def forward(
+        self,
+        src_tokens,
+        incremental_state: Optional[Dict[str, List[torch.Tensor]]] = None,
+        encoder_out=None,
+    ):
+        bsz = src_tokens.size(0)
+        if incremental_state is not None:  # used during inference
+            mems = self.get_incremental_state("mems")
+            src_tokens = src_tokens[:, -1:]  # only keep the most recent token
+        else:
+            mems = self._mems
+
+        if mems is None:
+            # first time init
+            mems = self.init_hid_cache(bsz)
+        output = self.model(x=src_tokens, h_cache=mems,)
+        if incremental_state is not None:
+            self.set_incremental_state(incremental_state, "mems", output[1])
+        else:
+            self._mems = output[1]
+        return (output[0],)
+
+    def max_positions(self):
+        return self.config.attn_span
+
+    def init_hid_cache(self, batch_sz):
+        hid = []
+        for layer in self.model.layers:
+            param = next(self.model.parameters())
+            h = torch.zeros(
+                batch_sz,
+                layer.get_cache_size(),
+                self.config.d_model,
+                dtype=param.dtype,
+                device=param.device,
+            )
+            hid.append(h)
+        return hid
+
+    def get_aux_loss(self):
+        return self.model.get_aux_loss()
+
+    def get_current_max_span(self):
+        return self.model.get_current_max_span()
+
+    def get_current_avg_span(self):
+        return self.model.get_current_avg_span()
+
+    def reorder_incremental_state(
+        self,
+        incremental_state: Dict[str, Dict[str, Optional[torch.Tensor]]],
+        new_order: torch.Tensor,
+    ):
+        """Reorder incremental state.
+
+        This will be called when the order of the input has changed from the
+        previous time step. A typical use case is beam search, where the input
+        order changes between time steps based on the selection of beams.
+        """
+        raise NotImplementedError("This is required for generation/beam search")
+        # mems = self.get_incremental_state(incremental_state, "mems")
+        # if mems is not None:
+        #     new_mems = [mems_i.index_select(1, new_order) for mems_i in mems]
+        #     self.set_incremental_state(incremental_state, "mems", new_mems)

fairseq/examples/adaptive_span/truncated_bptt_lm_task.py

@@ -0,0 +1,281 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+import os
+from dataclasses import dataclass, field
+from typing import List, Optional, Tuple
+
+import torch
+from fairseq import utils
+from fairseq.data import (
+    Dictionary,
+    TokenBlockDataset,
+    data_utils,
+    iterators,
+)
+from fairseq.dataclass import FairseqDataclass
+from fairseq.distributed import utils as dist_utils
+from fairseq.tasks import FairseqTask, register_task
+from omegaconf import II
+
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class TruncatedBPTTLMConfig(FairseqDataclass):
+    data: str = field(default="???", metadata={"help": "path to data directory"})
+    tokens_per_sample: int = field(
+        default=1024,
+        metadata={"help": "max number of tokens per sequence"},
+    )
+    batch_size: int = II("dataset.batch_size")
+    # Some models use *max_target_positions* to know how many positional
+    # embeddings to learn. We use II(...) to make it default to
+    # *tokens_per_sample*, but in principle there could be more positional
+    # embeddings than tokens in a single batch. This may also be irrelevant for
+    # custom model implementations.
+    max_target_positions: int = II("task.tokens_per_sample")
+    # these will be populated automatically if not provided
+    data_parallel_rank: Optional[int] = None
+    data_parallel_size: Optional[int] = None
+
+
+@register_task("truncated_bptt_lm", dataclass=TruncatedBPTTLMConfig)
+class TruncatedBPTTLMTask(FairseqTask):
+    def __init__(self, cfg: TruncatedBPTTLMConfig):
+        super().__init__(cfg)
+
+        if cfg.data_parallel_rank is None or cfg.data_parallel_size is None:
+            if torch.distributed.is_initialized():
+                cfg.data_parallel_rank = dist_utils.get_data_parallel_rank()
+                cfg.data_parallel_size = dist_utils.get_data_parallel_world_size()
+            else:
+                cfg.data_parallel_rank = 0
+                cfg.data_parallel_size = 1
+
+        # load the dictionary
+        paths = utils.split_paths(cfg.data)
+        assert len(paths) > 0
+        self.dictionary = Dictionary.load(os.path.join(paths[0], "dict.txt"))
+        logger.info("dictionary: {} types".format(len(self.dictionary)))
+
+    def load_dataset(self, split, epoch=1, combine=False, **kwargs):
+        """Load a given dataset split (e.g., train, valid, test)"""
+
+        # support sharded datasets
+        paths = utils.split_paths(self.cfg.data)
+        assert len(paths) > 0
+        data_path = paths[(epoch - 1) % len(paths)]
+        split_path = os.path.join(data_path, split)
+
+        # each element of *data* will be a tensorized line from the original
+        # text dataset, similar to ``open(split_path).readlines()``
+        data = data_utils.load_indexed_dataset(
+            split_path, self.dictionary, combine=combine
+        )
+        if data is None:
+            raise FileNotFoundError(
+                "Dataset not found: {} ({})".format(split, split_path)
+            )
+
+        # this is similar to ``data.view(-1).split(tokens_per_sample)``
+        data = TokenBlockDataset(
+            data,
+            data.sizes,
+            block_size=self.cfg.tokens_per_sample,
+            pad=None,  # unused
+            eos=None,  # unused
+            break_mode="none",
+        )
+
+        self.datasets[split] = TruncatedBPTTDataset(
+            data=data,
+            bsz_per_shard=self.cfg.batch_size,
+            shard_id=self.cfg.data_parallel_rank,
+            num_shards=self.cfg.data_parallel_size,
+        )
+
+    def dataset(self, split):
+        return self.datasets[split]
+
+    def get_batch_iterator(
+        self, dataset, num_workers=0, epoch=1, data_buffer_size=0, **kwargs
+    ):
+        return iterators.EpochBatchIterator(
+            dataset=dataset,
+            collate_fn=self._collate_fn,
+            num_workers=num_workers,
+            epoch=epoch,
+            buffer_size=data_buffer_size,
+            # we don't use the batching functionality from EpochBatchIterator;
+            # instead every item in *dataset* is a whole batch
+            batch_sampler=[[i] for i in range(len(dataset))],
+            disable_shuffling=True,
+        )
+
+    def _collate_fn(self, items: List[List[torch.Tensor]]):
+        # we don't use fairseq's batching functionality, so we expect a single
+        # Tensor of type List[torch.Tensor]
+        assert len(items) == 1
+
+        # item will have shape B x T (the last batch may have length < T)
+        id, item = items[0]
+        item = data_utils.collate_tokens(item, pad_idx=self.source_dictionary.pad())
+        B, T = item.size()
+
+        # shift item one position over and append a padding token for the target
+        target = torch.nn.functional.pad(
+            item[:, 1:], (0, 1, 0, 0), value=self.target_dictionary.pad()
+        )
+
+        # fairseq expects batches to have the following structure
+        return {
+            "id": torch.tensor([id]*item.size(0)),
+            "net_input": {
+                "src_tokens": item,
+            },
+            "target": target,
+            "nsentences": item.size(0),
+            "ntokens": item.numel(),
+        }
+
+    def build_dataset_for_inference(
+        self, src_tokens: List[torch.Tensor], src_lengths: List[int], **kwargs
+    ) -> torch.utils.data.Dataset:
+        eos = self.source_dictionary.eos()
+        dataset = TokenBlockDataset(
+            src_tokens,
+            src_lengths,
+            block_size=None,  # ignored for "eos" break mode
+            pad=self.source_dictionary.pad(),
+            eos=eos,
+            break_mode="eos",
+        )
+
+        class Dataset(torch.utils.data.Dataset):
+            def __getitem__(self, i):
+                item = dataset[i]
+                if item[-1] == eos:
+                    # remove eos to support generating with a prefix
+                    item = item[:-1]
+                return (i, [item])
+
+            def __len__(self):
+                return len(dataset)
+
+        return Dataset()
+
+    def inference_step(
+        self, generator, models, sample, prefix_tokens=None, constraints=None
+    ):
+        with torch.no_grad():
+            if constraints is not None:
+                raise NotImplementedError
+
+            # SequenceGenerator doesn't use *src_tokens* directly, we need to
+            # pass the *prefix_tokens* argument instead.
+            if prefix_tokens is None and sample["net_input"]["src_tokens"].nelement():
+                prefix_tokens = sample["net_input"]["src_tokens"]
+
+            # begin generation with the end-of-sentence token
+            bos_token = self.source_dictionary.eos()
+
+            return generator.generate(
+                models, sample, prefix_tokens=prefix_tokens, bos_token=bos_token
+            )
+
+    def eval_lm_dataloader(
+        self,
+        dataset,
+        max_tokens: Optional[int] = 36000,
+        batch_size: Optional[int] = None,
+        max_positions: Optional[int] = None,
+        num_shards: int = 1,
+        shard_id: int = 0,
+        num_workers: int = 1,
+        data_buffer_size: int = 10,
+        context_window: int = 0,
+    ):
+        if context_window > 0:
+            raise NotImplementedError(
+                "Transformer-XL doesn't need --context-window, try "
+                "--model-overrides '{\"mem_len\":42}' instead "
+            )
+        return self.get_batch_iterator(
+            dataset=dataset,
+            max_tokens=max_tokens,
+            max_sentences=batch_size,
+            max_positions=max_positions,
+            ignore_invalid_inputs=True,
+            num_shards=num_shards,
+            shard_id=shard_id,
+            num_workers=num_workers,
+            data_buffer_size=data_buffer_size,
+        ).next_epoch_itr(shuffle=False)
+
+    @property
+    def source_dictionary(self):
+        return self.dictionary
+
+    @property
+    def target_dictionary(self):
+        return self.dictionary
+
+
+class TruncatedBPTTDataset(torch.utils.data.Dataset):
+    def __init__(
+        self,
+        data: List[torch.Tensor],  # ordered list of items
+        bsz_per_shard,  # number of items processed per GPUs per forward
+        shard_id,  # current GPU ID
+        num_shards,  # number of GPUs
+    ):
+        super().__init__()
+        self.data = data
+
+        def batchify(data, bsz):
+            # Work out how cleanly we can divide the dataset into bsz parts.
+            nbatch = data.size(0) // bsz
+            # Trim off any extra elements that wouldn't cleanly fit (remainders).
+            data = data.narrow(0, 0, nbatch * bsz)
+            # Evenly divide the data across the bsz batches.
+            data = data.view(bsz, -1).contiguous()
+            return data
+
+        # total number of sequences processed by all GPUs in each forward pass
+        global_batch_size = bsz_per_shard * num_shards
+
+        """
+        With a 16 item dataset, bsz_per_shard=2 and num_shards=3,
+        *indices* might look like:
+
+            indices = [[0, 1],
+                       [2, 3],
+                       [4, 5],
+                       [6, 7],
+                       [8, 9],
+                       [10, 11]]
+
+        The size of the TruncatedBPTTDataset instance will be 2,
+        and shard 1 will see items:
+
+            [(0, [data[4], data[6]]),
+             (1, [data[5], data[7]])]
+        """
+        indices = batchify(torch.arange(len(data)), global_batch_size)
+        assert indices.size(0) == global_batch_size
+
+        self.my_indices = indices[
+            shard_id * bsz_per_shard : (shard_id + 1) * bsz_per_shard
+        ]
+        assert self.my_indices.size(0) == bsz_per_shard
+
+    def __len__(self):
+        return self.my_indices.size(1)
+
+    def __getitem__(self, i) -> Tuple[int, List[torch.Tensor]]:
+        return (i, [self.data[idx] for idx in self.my_indices[:, i]])

fairseq/examples/backtranslation/README.md

@@ -0,0 +1,297 @@
+# Understanding Back-Translation at Scale (Edunov et al., 2018)
+
+This page includes pre-trained models from the paper [Understanding Back-Translation at Scale (Edunov et al., 2018)](https://arxiv.org/abs/1808.09381).
+
+## Pre-trained models
+
+Model | Description | Dataset | Download
+---|---|---|---
+`transformer.wmt18.en-de` | Transformer <br> ([Edunov et al., 2018](https://arxiv.org/abs/1808.09381)) <br> WMT'18 winner | [WMT'18 English-German](http://www.statmt.org/wmt18/translation-task.html) | [download (.tar.gz)](https://dl.fbaipublicfiles.com/fairseq/models/wmt18.en-de.ensemble.tar.gz) <br> See NOTE in the archive
+
+## Example usage (torch.hub)
+
+We require a few additional Python dependencies for preprocessing:
+```bash
+pip install subword_nmt sacremoses
+```
+
+Then to generate translations from the full model ensemble:
+```python
+import torch
+
+# List available models
+torch.hub.list('pytorch/fairseq')  # [..., 'transformer.wmt18.en-de', ... ]
+
+# Load the WMT'18 En-De ensemble
+en2de_ensemble = torch.hub.load(
+    'pytorch/fairseq', 'transformer.wmt18.en-de',
+    checkpoint_file='wmt18.model1.pt:wmt18.model2.pt:wmt18.model3.pt:wmt18.model4.pt:wmt18.model5.pt',
+    tokenizer='moses', bpe='subword_nmt')
+
+# The ensemble contains 5 models
+len(en2de_ensemble.models)
+# 5
+
+# Translate
+en2de_ensemble.translate('Hello world!')
+# 'Hallo Welt!'
+```
+
+## Training your own model (WMT'18 English-German)
+
+The following instructions can be adapted to reproduce the models from the paper.
+
+
+#### Step 1. Prepare parallel data and optionally train a baseline (English-German) model
+
+First download and preprocess the data:
+```bash
+# Download and prepare the data
+cd examples/backtranslation/
+bash prepare-wmt18en2de.sh
+cd ../..
+
+# Binarize the data
+TEXT=examples/backtranslation/wmt18_en_de
+fairseq-preprocess \
+    --joined-dictionary \
+    --source-lang en --target-lang de \
+    --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
+    --destdir data-bin/wmt18_en_de --thresholdtgt 0 --thresholdsrc 0 \
+    --workers 20
+
+# Copy the BPE code into the data-bin directory for future use
+cp examples/backtranslation/wmt18_en_de/code data-bin/wmt18_en_de/code
+```
+
+(Optionally) Train a baseline model (English-German) using just the parallel data:
+```bash
+CHECKPOINT_DIR=checkpoints_en_de_parallel
+fairseq-train --fp16 \
+    data-bin/wmt18_en_de \
+    --source-lang en --target-lang de \
+    --arch transformer_wmt_en_de_big --share-all-embeddings \
+    --dropout 0.3 --weight-decay 0.0 \
+    --criterion label_smoothed_cross_entropy --label-smoothing 0.1 \
+    --optimizer adam --adam-betas '(0.9, 0.98)' --clip-norm 0.0 \
+    --lr 0.001 --lr-scheduler inverse_sqrt --warmup-updates 4000 \
+    --max-tokens 3584 --update-freq 16 \
+    --max-update 30000 \
+    --save-dir $CHECKPOINT_DIR
+# Note: the above command assumes 8 GPUs. Adjust `--update-freq` if you have a
+# different number of GPUs.
+```
+
+Average the last 10 checkpoints:
+```bash
+python scripts/average_checkpoints.py \
+    --inputs $CHECKPOINT_DIR \
+    --num-epoch-checkpoints 10 \
+    --output $CHECKPOINT_DIR/checkpoint.avg10.pt
+```
+
+Evaluate BLEU:
+```bash
+# tokenized BLEU on newstest2017:
+bash examples/backtranslation/tokenized_bleu.sh \
+    wmt17 \
+    en-de \
+    data-bin/wmt18_en_de \
+    data-bin/wmt18_en_de/code \
+    $CHECKPOINT_DIR/checkpoint.avg10.pt
+# BLEU4 = 29.57, 60.9/35.4/22.9/15.5 (BP=1.000, ratio=1.014, syslen=63049, reflen=62152)
+# compare to 29.46 in Table 1, which is also for tokenized BLEU
+
+# generally it's better to report (detokenized) sacrebleu though:
+bash examples/backtranslation/sacrebleu.sh \
+    wmt17 \
+    en-de \
+    data-bin/wmt18_en_de \
+    data-bin/wmt18_en_de/code \
+    $CHECKPOINT_DIR/checkpoint.avg10.pt
+# BLEU+case.mixed+lang.en-de+numrefs.1+smooth.exp+test.wmt17+tok.13a+version.1.4.3 = 29.0 60.6/34.7/22.4/14.9 (BP = 1.000 ratio = 1.013 hyp_len = 62099 ref_len = 61287)
+```
+
+
+#### Step 2. Back-translate monolingual German data
+
+Train a reverse model (German-English) to do the back-translation:
+```bash
+CHECKPOINT_DIR=checkpoints_de_en_parallel
+fairseq-train --fp16 \
+    data-bin/wmt18_en_de \
+    --source-lang de --target-lang en \
+    --arch transformer_wmt_en_de_big --share-all-embeddings \
+    --dropout 0.3 --weight-decay 0.0 \
+    --criterion label_smoothed_cross_entropy --label-smoothing 0.1 \
+    --optimizer adam --adam-betas '(0.9, 0.98)' --clip-norm 0.0 \
+    --lr 0.001 --lr-scheduler inverse_sqrt --warmup-updates 4000 \
+    --max-tokens 3584 --update-freq 16 \
+    --max-update 30000 \
+    --save-dir $CHECKPOINT_DIR
+# Note: the above command assumes 8 GPUs. Adjust `--update-freq` if you have a
+# different number of GPUs.
+```
+
+Let's evaluate the back-translation (BT) model to make sure it is well trained:
+```bash
+bash examples/backtranslation/sacrebleu.sh \
+    wmt17 \
+    de-en \
+    data-bin/wmt18_en_de \
+    data-bin/wmt18_en_de/code \
+    $CHECKPOINT_DIR/checkpoint_best.py
+# BLEU+case.mixed+lang.de-en+numrefs.1+smooth.exp+test.wmt17+tok.13a+version.1.4.3 = 34.9 66.9/41.8/28.5/19.9 (BP = 0.983 ratio = 0.984 hyp_len = 63342 ref_len = 64399)
+# compare to the best system from WMT'17 which scored 35.1: http://matrix.statmt.org/matrix/systems_list/1868
+```
+
+Next prepare the monolingual data:
+```bash
+# Download and prepare the monolingual data
+# By default the script samples 25M monolingual sentences, which after
+# deduplication should be just over 24M sentences. These are split into 25
+# shards, each with 1M sentences (except for the last shard).
+cd examples/backtranslation/
+bash prepare-de-monolingual.sh
+cd ../..
+
+# Binarize each shard of the monolingual data
+TEXT=examples/backtranslation/wmt18_de_mono
+for SHARD in $(seq -f "%02g" 0 24); do \
+    fairseq-preprocess \
+        --only-source \
+        --source-lang de --target-lang en \
+        --joined-dictionary \
+        --srcdict data-bin/wmt18_en_de/dict.de.txt \
+        --testpref $TEXT/bpe.monolingual.dedup.${SHARD} \
+        --destdir data-bin/wmt18_de_mono/shard${SHARD} \
+        --workers 20; \
+    cp data-bin/wmt18_en_de/dict.en.txt data-bin/wmt18_de_mono/shard${SHARD}/; \
+done
+```
+
+Now we're ready to perform back-translation over the monolingual data. The
+following command generates via sampling, but it's possible to use greedy
+decoding (`--beam 1`), beam search (`--beam 5`),
+top-k sampling (`--sampling --beam 1 --sampling-topk 10`), etc.:
+```bash
+mkdir backtranslation_output
+for SHARD in $(seq -f "%02g" 0 24); do \
+    fairseq-generate --fp16 \
+        data-bin/wmt18_de_mono/shard${SHARD} \
+        --path $CHECKPOINT_DIR/checkpoint_best.pt \
+        --skip-invalid-size-inputs-valid-test \
+        --max-tokens 4096 \
+        --sampling --beam 1 \
+    > backtranslation_output/sampling.shard${SHARD}.out; \
+done
+```
+
+After BT, use the `extract_bt_data.py` script to re-combine the shards, extract
+the back-translations and apply length ratio filters:
+```bash
+python examples/backtranslation/extract_bt_data.py \
+    --minlen 1 --maxlen 250 --ratio 1.5 \
+    --output backtranslation_output/bt_data --srclang en --tgtlang de \
+    backtranslation_output/sampling.shard*.out
+
+# Ensure lengths are the same:
+# wc -l backtranslation_output/bt_data.{en,de}
+#   21795614 backtranslation_output/bt_data.en
+#   21795614 backtranslation_output/bt_data.de
+#   43591228 total
+```
+
+Binarize the filtered BT data and combine it with the parallel data:
+```bash
+TEXT=backtranslation_output
+fairseq-preprocess \
+    --source-lang en --target-lang de \
+    --joined-dictionary \
+    --srcdict data-bin/wmt18_en_de/dict.en.txt \
+    --trainpref $TEXT/bt_data \
+    --destdir data-bin/wmt18_en_de_bt \
+    --workers 20
+
+# We want to train on the combined data, so we'll symlink the parallel + BT data
+# in the wmt18_en_de_para_plus_bt directory. We link the parallel data as "train"
+# and the BT data as "train1", so that fairseq will combine them automatically
+# and so that we can use the `--upsample-primary` option to upsample the
+# parallel data (if desired).
+PARA_DATA=$(readlink -f data-bin/wmt18_en_de)
+BT_DATA=$(readlink -f data-bin/wmt18_en_de_bt)
+COMB_DATA=data-bin/wmt18_en_de_para_plus_bt
+mkdir -p $COMB_DATA
+for LANG in en de; do \
+    ln -s ${PARA_DATA}/dict.$LANG.txt ${COMB_DATA}/dict.$LANG.txt; \
+    for EXT in bin idx; do \
+        ln -s ${PARA_DATA}/train.en-de.$LANG.$EXT ${COMB_DATA}/train.en-de.$LANG.$EXT; \
+        ln -s ${BT_DATA}/train.en-de.$LANG.$EXT ${COMB_DATA}/train1.en-de.$LANG.$EXT; \
+        ln -s ${PARA_DATA}/valid.en-de.$LANG.$EXT ${COMB_DATA}/valid.en-de.$LANG.$EXT; \
+        ln -s ${PARA_DATA}/test.en-de.$LANG.$EXT ${COMB_DATA}/test.en-de.$LANG.$EXT; \
+    done; \
+done
+```
+
+
+#### 3. Train an English-German model over the combined parallel + BT data
+
+Finally we can train a model over the parallel + BT data:
+```bash
+CHECKPOINT_DIR=checkpoints_en_de_parallel_plus_bt
+fairseq-train --fp16 \
+    data-bin/wmt18_en_de_para_plus_bt \
+    --upsample-primary 16 \
+    --source-lang en --target-lang de \
+    --arch transformer_wmt_en_de_big --share-all-embeddings \
+    --dropout 0.3 --weight-decay 0.0 \
+    --criterion label_smoothed_cross_entropy --label-smoothing 0.1 \
+    --optimizer adam --adam-betas '(0.9, 0.98)' --clip-norm 0.0 \
+    --lr 0.0007 --lr-scheduler inverse_sqrt --warmup-updates 4000 \
+    --max-tokens 3584 --update-freq 16 \
+    --max-update 100000 \
+    --save-dir $CHECKPOINT_DIR
+# Note: the above command assumes 8 GPUs. Adjust `--update-freq` if you have a
+# different number of GPUs.
+```
+
+Average the last 10 checkpoints:
+```bash
+python scripts/average_checkpoints.py \
+    --inputs $CHECKPOINT_DIR \
+    --num-epoch-checkpoints 10 \
+    --output $CHECKPOINT_DIR/checkpoint.avg10.pt
+```
+
+Evaluate BLEU:
+```bash
+# tokenized BLEU on newstest2017:
+bash examples/backtranslation/tokenized_bleu.sh \
+    wmt17 \
+    en-de \
+    data-bin/wmt18_en_de \
+    data-bin/wmt18_en_de/code \
+    $CHECKPOINT_DIR/checkpoint.avg10.pt
+# BLEU4 = 32.35, 64.4/38.9/26.2/18.3 (BP=0.977, ratio=0.977, syslen=60729, reflen=62152)
+# compare to 32.35 in Table 1, which is also for tokenized BLEU
+
+# generally it's better to report (detokenized) sacrebleu:
+bash examples/backtranslation/sacrebleu.sh \
+    wmt17 \
+    en-de \
+    data-bin/wmt18_en_de \
+    data-bin/wmt18_en_de/code \
+    $CHECKPOINT_DIR/checkpoint.avg10.pt
+# BLEU+case.mixed+lang.en-de+numrefs.1+smooth.exp+test.wmt17+tok.13a+version.1.4.3 = 31.5 64.3/38.2/25.6/17.6 (BP = 0.971 ratio = 0.971 hyp_len = 59515 ref_len = 61287)
+```
+
+
+## Citation
+```bibtex
+@inproceedings{edunov2018backtranslation,
+  title = {Understanding Back-Translation at Scale},
+  author = {Edunov, Sergey and Ott, Myle and Auli, Michael and Grangier, David},
+  booktitle = {Conference of the Association for Computational Linguistics (ACL)},
+  year = 2018,
+}
+```

fairseq/examples/backtranslation/deduplicate_lines.py

@@ -0,0 +1,41 @@
+#!/usr/bin/python3
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import argparse
+import fileinput
+import hashlib
+import sys
+from multiprocessing import Pool
+
+
+def get_hashes_and_lines(raw_line):
+    hash = hashlib.md5(raw_line).hexdigest()
+    return hash, raw_line
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--workers", type=int, default=10)
+    parser.add_argument("files", nargs="*", help="input files")
+    args = parser.parse_args()
+
+    seen = set()
+    with fileinput.input(args.files, mode="rb") as h:
+        pool = Pool(args.workers)
+        results = pool.imap_unordered(get_hashes_and_lines, h, 1000)
+        for i, (hash, raw_line) in enumerate(results):
+            if hash not in seen:
+                seen.add(hash)
+                sys.stdout.buffer.write(raw_line)
+            if i % 1000000 == 0:
+                print(i, file=sys.stderr, end="", flush=True)
+            elif i % 100000 == 0:
+                print(".", file=sys.stderr, end="", flush=True)
+    print(file=sys.stderr, flush=True)
+
+
+if __name__ == "__main__":
+    main()

fairseq/examples/backtranslation/extract_bt_data.py

@@ -0,0 +1,72 @@
+#!/usr/bin/env python
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import argparse
+import fileinput
+
+from tqdm import tqdm
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description=(
+            "Extract back-translations from the stdout of fairseq-generate. "
+            "If there are multiply hypotheses for a source, we only keep the first one. "
+        )
+    )
+    parser.add_argument("--output", required=True, help="output prefix")
+    parser.add_argument(
+        "--srclang", required=True, help="source language (extracted from H-* lines)"
+    )
+    parser.add_argument(
+        "--tgtlang", required=True, help="target language (extracted from S-* lines)"
+    )
+    parser.add_argument("--minlen", type=int, help="min length filter")
+    parser.add_argument("--maxlen", type=int, help="max length filter")
+    parser.add_argument("--ratio", type=float, help="ratio filter")
+    parser.add_argument("files", nargs="*", help="input files")
+    args = parser.parse_args()
+
+    def validate(src, tgt):
+        srclen = len(src.split(" ")) if src != "" else 0
+        tgtlen = len(tgt.split(" ")) if tgt != "" else 0
+        if (
+            (args.minlen is not None and (srclen < args.minlen or tgtlen < args.minlen))
+            or (
+                args.maxlen is not None
+                and (srclen > args.maxlen or tgtlen > args.maxlen)
+            )
+            or (
+                args.ratio is not None
+                and (max(srclen, tgtlen) / float(min(srclen, tgtlen)) > args.ratio)
+            )
+        ):
+            return False
+        return True
+
+    def safe_index(toks, index, default):
+        try:
+            return toks[index]
+        except IndexError:
+            return default
+
+    with open(args.output + "." + args.srclang, "w") as src_h, open(
+        args.output + "." + args.tgtlang, "w"
+    ) as tgt_h:
+        for line in tqdm(fileinput.input(args.files)):
+            if line.startswith("S-"):
+                tgt = safe_index(line.rstrip().split("\t"), 1, "")
+            elif line.startswith("H-"):
+                if tgt is not None:
+                    src = safe_index(line.rstrip().split("\t"), 2, "")
+                    if validate(src, tgt):
+                        print(src, file=src_h)
+                        print(tgt, file=tgt_h)
+                    tgt = None
+
+
+if __name__ == "__main__":
+    main()

fairseq/examples/backtranslation/prepare-de-monolingual.sh

@@ -0,0 +1,98 @@
+#!/bin/bash
+
+SCRIPTS=mosesdecoder/scripts
+TOKENIZER=$SCRIPTS/tokenizer/tokenizer.perl
+NORM_PUNC=$SCRIPTS/tokenizer/normalize-punctuation.perl
+REM_NON_PRINT_CHAR=$SCRIPTS/tokenizer/remove-non-printing-char.perl
+BPEROOT=subword-nmt/subword_nmt
+
+
+BPE_CODE=wmt18_en_de/code
+SUBSAMPLE_SIZE=25000000
+LANG=de
+
+
+OUTDIR=wmt18_${LANG}_mono
+orig=orig
+tmp=$OUTDIR/tmp
+mkdir -p $OUTDIR $tmp
+
+
+URLS=(
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2007.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2008.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2009.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2010.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2011.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2012.de.shuffled.gz"
+    "http://www.statmt.org/wmt14/training-monolingual-news-crawl/news.2013.de.shuffled.gz"
+    "http://www.statmt.org/wmt15/training-monolingual-news-crawl-v2/news.2014.de.shuffled.v2.gz"
+    "http://data.statmt.org/wmt16/translation-task/news.2015.de.shuffled.gz"
+    "http://data.statmt.org/wmt17/translation-task/news.2016.de.shuffled.gz"
+    "http://data.statmt.org/wmt18/translation-task/news.2017.de.shuffled.deduped.gz"
+)
+FILES=(
+    "news.2007.de.shuffled.gz"
+    "news.2008.de.shuffled.gz"
+    "news.2009.de.shuffled.gz"
+    "news.2010.de.shuffled.gz"
+    "news.2011.de.shuffled.gz"
+    "news.2012.de.shuffled.gz"
+    "news.2013.de.shuffled.gz"
+    "news.2014.de.shuffled.v2.gz"
+    "news.2015.de.shuffled.gz"
+    "news.2016.de.shuffled.gz"
+    "news.2017.de.shuffled.deduped.gz"
+)
+
+
+cd $orig
+for ((i=0;i<${#URLS[@]};++i)); do
+    file=${FILES[i]}
+    if [ -f $file ]; then
+        echo "$file already exists, skipping download"
+    else
+        url=${URLS[i]}
+        wget "$url"
+    fi
+done
+cd ..
+
+
+if [ -f $tmp/monolingual.${SUBSAMPLE_SIZE}.${LANG} ]; then
+    echo "found monolingual sample, skipping shuffle/sample/tokenize"
+else
+    gzip -c -d -k $(for FILE in "${FILES[@]}"; do echo $orig/$FILE; done) \
+    | shuf -n $SUBSAMPLE_SIZE \
+    | perl $NORM_PUNC $LANG \
+    | perl $REM_NON_PRINT_CHAR \
+    | perl $TOKENIZER -threads 8 -a -l $LANG \
+    > $tmp/monolingual.${SUBSAMPLE_SIZE}.${LANG}
+fi
+
+
+if [ -f $tmp/bpe.monolingual.${SUBSAMPLE_SIZE}.${LANG} ]; then
+    echo "found BPE monolingual sample, skipping BPE step"
+else
+    python $BPEROOT/apply_bpe.py -c $BPE_CODE \
+        < $tmp/monolingual.${SUBSAMPLE_SIZE}.${LANG} \
+        > $tmp/bpe.monolingual.${SUBSAMPLE_SIZE}.${LANG}
+fi
+
+
+if [ -f $tmp/bpe.monolingual.dedup.${SUBSAMPLE_SIZE}.${LANG} ]; then
+    echo "found deduplicated monolingual sample, skipping deduplication step"
+else
+    python deduplicate_lines.py $tmp/bpe.monolingual.${SUBSAMPLE_SIZE}.${LANG} \
+    > $tmp/bpe.monolingual.dedup.${SUBSAMPLE_SIZE}.${LANG}
+fi
+
+
+if [ -f $OUTDIR/bpe.monolingual.dedup.00.de ]; then
+    echo "found sharded data, skipping sharding step"
+else
+    split --lines 1000000 --numeric-suffixes \
+        --additional-suffix .${LANG} \
+        $tmp/bpe.monolingual.dedup.${SUBSAMPLE_SIZE}.${LANG} \
+        $OUTDIR/bpe.monolingual.dedup.
+fi

fairseq/examples/backtranslation/prepare-wmt18en2de.sh

@@ -0,0 +1,135 @@
+#!/bin/bash
+# Adapted from https://github.com/facebookresearch/MIXER/blob/master/prepareData.sh
+
+echo 'Cloning Moses github repository (for tokenization scripts)...'
+git clone https://github.com/moses-smt/mosesdecoder.git
+
+echo 'Cloning Subword NMT repository (for BPE pre-processing)...'
+git clone https://github.com/rsennrich/subword-nmt.git
+
+SCRIPTS=mosesdecoder/scripts
+TOKENIZER=$SCRIPTS/tokenizer/tokenizer.perl
+CLEAN=$SCRIPTS/training/clean-corpus-n.perl
+NORM_PUNC=$SCRIPTS/tokenizer/normalize-punctuation.perl
+REM_NON_PRINT_CHAR=$SCRIPTS/tokenizer/remove-non-printing-char.perl
+BPEROOT=subword-nmt/subword_nmt
+BPE_TOKENS=32000
+
+URLS=(
+    "http://statmt.org/wmt13/training-parallel-europarl-v7.tgz"
+    "http://statmt.org/wmt13/training-parallel-commoncrawl.tgz"
+    "http://data.statmt.org/wmt18/translation-task/training-parallel-nc-v13.tgz"
+    "http://data.statmt.org/wmt18/translation-task/rapid2016.tgz"
+    "http://data.statmt.org/wmt17/translation-task/dev.tgz"
+    "http://statmt.org/wmt14/test-full.tgz"
+)
+FILES=(
+    "training-parallel-europarl-v7.tgz"
+    "training-parallel-commoncrawl.tgz"
+    "training-parallel-nc-v13.tgz"
+    "rapid2016.tgz"
+    "dev.tgz"
+    "test-full.tgz"
+)
+CORPORA=(
+    "training/europarl-v7.de-en"
+    "commoncrawl.de-en"
+    "training-parallel-nc-v13/news-commentary-v13.de-en"
+    "rapid2016.de-en"
+)
+
+if [ ! -d "$SCRIPTS" ]; then
+    echo "Please set SCRIPTS variable correctly to point to Moses scripts."
+    exit 1
+fi
+
+OUTDIR=wmt18_en_de
+
+src=en
+tgt=de
+lang=en-de
+prep=$OUTDIR
+tmp=$prep/tmp
+orig=orig
+
+mkdir -p $orig $tmp $prep
+
+cd $orig
+
+for ((i=0;i<${#URLS[@]};++i)); do
+    file=${FILES[i]}
+    if [ -f $file ]; then
+        echo "$file already exists, skipping download"
+    else
+        url=${URLS[i]}
+        wget "$url"
+        if [ -f $file ]; then
+            echo "$url successfully downloaded."
+        else
+            echo "$url not successfully downloaded."
+            exit 1
+        fi
+        if [ ${file: -4} == ".tgz" ]; then
+            tar zxvf $file
+        elif [ ${file: -4} == ".tar" ]; then
+            tar xvf $file
+        fi
+    fi
+done
+cd ..
+
+echo "pre-processing train data..."
+for l in $src $tgt; do
+    rm $tmp/train.tags.$lang.tok.$l
+    for f in "${CORPORA[@]}"; do
+        cat $orig/$f.$l | \
+            perl $NORM_PUNC $l | \
+            perl $REM_NON_PRINT_CHAR | \
+            perl $TOKENIZER -threads 8 -a -l $l >> $tmp/train.tags.$lang.tok.$l
+    done
+done
+
+echo "pre-processing test data..."
+for l in $src $tgt; do
+    if [ "$l" == "$src" ]; then
+        t="src"
+    else
+        t="ref"
+    fi
+    grep '<seg id' $orig/test-full/newstest2014-deen-$t.$l.sgm | \
+        sed -e 's/<seg id="[0-9]*">\s*//g' | \
+        sed -e 's/\s*<\/seg>\s*//g' | \
+        sed -e "s/\’/\'/g" | \
+    perl $TOKENIZER -threads 8 -a -l $l > $tmp/test.$l
+    echo ""
+done
+
+echo "splitting train and valid..."
+for l in $src $tgt; do
+    awk '{if (NR%100 == 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/valid.$l
+    awk '{if (NR%100 != 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/train.$l
+done
+
+TRAIN=$tmp/train.de-en
+BPE_CODE=$prep/code
+rm -f $TRAIN
+for l in $src $tgt; do
+    cat $tmp/train.$l >> $TRAIN
+done
+
+echo "learn_bpe.py on ${TRAIN}..."
+python $BPEROOT/learn_bpe.py -s $BPE_TOKENS < $TRAIN > $BPE_CODE
+
+for L in $src $tgt; do
+    for f in train.$L valid.$L test.$L; do
+        echo "apply_bpe.py to ${f}..."
+        python $BPEROOT/apply_bpe.py -c $BPE_CODE < $tmp/$f > $tmp/bpe.$f
+    done
+done
+
+perl $CLEAN -ratio 1.5 $tmp/bpe.train $src $tgt $prep/train 1 250
+perl $CLEAN -ratio 1.5 $tmp/bpe.valid $src $tgt $prep/valid 1 250
+
+for L in $src $tgt; do
+    cp $tmp/bpe.test.$L $prep/test.$L
+done

fairseq/examples/backtranslation/sacrebleu.sh

@@ -0,0 +1,37 @@
+#!/bin/bash
+
+if [ $# -ne 5 ]; then
+    echo "usage: $0 [dataset=wmt14/full] [langpair=en-de] [databin] [bpecode] [model]"
+    exit
+fi
+
+
+DATASET=$1
+LANGPAIR=$2
+DATABIN=$3
+BPECODE=$4
+MODEL=$5
+
+SRCLANG=$(echo $LANGPAIR | cut -d '-' -f 1)
+TGTLANG=$(echo $LANGPAIR | cut -d '-' -f 2)
+
+
+BPEROOT=examples/backtranslation/subword-nmt/subword_nmt
+if [ ! -e $BPEROOT ]; then
+    BPEROOT=subword-nmt/subword_nmt
+    if [ ! -e $BPEROOT ]; then
+        echo 'Cloning Subword NMT repository (for BPE pre-processing)...'
+        git clone https://github.com/rsennrich/subword-nmt.git
+    fi
+fi
+
+
+sacrebleu -t $DATASET -l $LANGPAIR --echo src \
+| sacremoses tokenize -a -l $SRCLANG -q \
+| python $BPEROOT/apply_bpe.py -c $BPECODE \
+| fairseq-interactive $DATABIN --path $MODEL \
+    -s $SRCLANG -t $TGTLANG \
+    --beam 5 --remove-bpe --buffer-size 1024 --max-tokens 8000 \
+| grep ^H- | cut -f 3- \
+| sacremoses detokenize -l $TGTLANG -q \
+| sacrebleu -t $DATASET -l $LANGPAIR

fairseq/examples/backtranslation/tokenized_bleu.sh

@@ -0,0 +1,46 @@
+#!/bin/bash
+
+if [ $# -ne 5 ]; then
+    echo "usage: $0 [dataset=wmt14/full] [langpair=en-de] [databin] [bpecode] [model]"
+    exit
+fi
+
+
+DATASET=$1
+LANGPAIR=$2
+DATABIN=$3
+BPECODE=$4
+MODEL=$5
+
+SRCLANG=$(echo $LANGPAIR | cut -d '-' -f 1)
+TGTLANG=$(echo $LANGPAIR | cut -d '-' -f 2)
+
+
+BPEROOT=examples/backtranslation/subword-nmt/subword_nmt
+if [ ! -e $BPEROOT ]; then
+    BPEROOT=subword-nmt/subword_nmt
+    if [ ! -e $BPEROOT ]; then
+        echo 'Cloning Subword NMT repository (for BPE pre-processing)...'
+        git clone https://github.com/rsennrich/subword-nmt.git
+    fi
+fi
+
+
+TMP_REF=$(mktemp)
+
+sacrebleu -t $DATASET -l $LANGPAIR --echo ref -q \
+| sacremoses normalize -l $TGTLANG -q \
+| sacremoses tokenize -a -l $TGTLANG -q \
+> $TMP_REF
+
+sacrebleu -t $DATASET -l $LANGPAIR --echo src -q \
+| sacremoses normalize -l $SRCLANG -q \
+| sacremoses tokenize -a -l $SRCLANG -q \
+| python $BPEROOT/apply_bpe.py -c $BPECODE \
+| fairseq-interactive $DATABIN --path $MODEL \
+    -s $SRCLANG -t $TGTLANG \
+    --beam 5 --remove-bpe --buffer-size 1024 --max-tokens 8000 \
+| grep ^H- | cut -f 3- \
+| fairseq-score --ref $TMP_REF
+
+rm -f $TMP_REF

fairseq/examples/bart/README.glue.md

@@ -0,0 +1,99 @@
+# Fine-tuning BART on GLUE tasks
+
+### 1) Download the data from GLUE website (https://gluebenchmark.com/tasks) using following commands:
+```bash
+wget https://gist.githubusercontent.com/W4ngatang/60c2bdb54d156a41194446737ce03e2e/raw/17b8dd0d724281ed7c3b2aeeda662b92809aadd5/download_glue_data.py
+python download_glue_data.py --data_dir glue_data --tasks all
+```
+
+### 2) Preprocess GLUE task data (same as RoBERTa):
+```bash
+./examples/roberta/preprocess_GLUE_tasks.sh glue_data <glue_task_name>
+```
+`glue_task_name` is one of the following:
+`{ALL, QQP, MNLI, QNLI, MRPC, RTE, STS-B, SST-2, CoLA}`
+Use `ALL` for preprocessing all the glue tasks.
+
+### 3) Fine-tuning on GLUE task:
+Example fine-tuning cmd for `RTE` task
+```bash
+TOTAL_NUM_UPDATES=2036  # 10 epochs through RTE for bsz 16
+WARMUP_UPDATES=61      # 6 percent of the number of updates
+LR=1e-05                # Peak LR for polynomial LR scheduler.
+NUM_CLASSES=2
+MAX_SENTENCES=16        # Batch size.
+BART_PATH=/path/to/bart/model.pt
+
+CUDA_VISIBLE_DEVICES=0,1 fairseq-train RTE-bin/ \
+    --restore-file $BART_PATH \
+    --batch-size $MAX_SENTENCES \
+    --max-tokens 4400 \
+    --task sentence_prediction \
+    --add-prev-output-tokens \
+    --layernorm-embedding \
+    --share-all-embeddings \
+    --share-decoder-input-output-embed \
+    --reset-optimizer --reset-dataloader --reset-meters \
+    --required-batch-size-multiple 1 \
+    --init-token 0 \
+    --arch bart_large \
+    --criterion sentence_prediction \
+    --num-classes $NUM_CLASSES \
+    --dropout 0.1 --attention-dropout 0.1 \
+    --weight-decay 0.01 --optimizer adam --adam-betas "(0.9, 0.98)" --adam-eps 1e-08 \
+    --clip-norm 0.0 \
+    --lr-scheduler polynomial_decay --lr $LR --total-num-update $TOTAL_NUM_UPDATES --warmup-updates $WARMUP_UPDATES \
+    --fp16 --fp16-init-scale 4 --threshold-loss-scale 1 --fp16-scale-window 128 \
+    --max-epoch 10 \
+    --find-unused-parameters \
+    --best-checkpoint-metric accuracy --maximize-best-checkpoint-metric;
+```
+
+For each of the GLUE task, you will need to use following cmd-line arguments:
+
+Model | MNLI | QNLI | QQP | RTE | SST-2 | MRPC | CoLA | STS-B
+---|---|---|---|---|---|---|---|---
+`--num-classes` | 3 | 2 | 2 | 2 | 2 | 2 | 2 | 1
+`--lr` | 5e-6 | 1e-5 | 1e-5 | 1e-5 | 5e-6 | 2e-5 | 2e-5 | 2e-5
+`bsz` | 128 | 32 | 32 | 32 | 128 | 64 | 64 | 32
+`--total-num-update` | 30968 | 33112 | 113272 | 1018 | 5233 | 1148 | 1334 | 1799
+`--warmup-updates` | 1858 | 1986 | 6796 | 61 | 314 | 68 | 80 | 107
+
+For `STS-B` additionally add `--regression-target --best-checkpoint-metric loss` and remove `--maximize-best-checkpoint-metric`.
+
+**Note:**
+
+a) `--total-num-updates` is used by `--polynomial_decay` scheduler and is calculated for `--max-epoch=10` and `--batch-size=32/64/128` depending on the task.
+
+b) Above cmd-args and hyperparams are tested on Nvidia `V100` GPU with `32gb` of memory for each task. Depending on the GPU memory resources available to you, you can use increase `--update-freq` and reduce `--batch-size`.
+
+### Inference on GLUE task
+After training the model as mentioned in previous step, you can perform inference with checkpoints in `checkpoints/` directory using following python code snippet:
+
+```python
+from fairseq.models.bart import BARTModel
+
+bart = BARTModel.from_pretrained(
+    'checkpoints/',
+    checkpoint_file='checkpoint_best.pt',
+    data_name_or_path='RTE-bin'
+)
+
+label_fn = lambda label: bart.task.label_dictionary.string(
+    [label + bart.task.label_dictionary.nspecial]
+)   
+ncorrect, nsamples = 0, 0
+bart.cuda()
+bart.eval()
+with open('glue_data/RTE/dev.tsv') as fin:
+    fin.readline()
+    for index, line in enumerate(fin):
+        tokens = line.strip().split('\t')
+        sent1, sent2, target = tokens[1], tokens[2], tokens[3]
+        tokens = bart.encode(sent1, sent2)
+        prediction = bart.predict('sentence_classification_head', tokens).argmax().item()
+        prediction_label = label_fn(prediction)
+        ncorrect += int(prediction_label == target)
+        nsamples += 1
+print('| Accuracy: ', float(ncorrect)/float(nsamples))
+```

fairseq/examples/bart/README.md

@@ -0,0 +1,228 @@
+# BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension
+
+[https://arxiv.org/abs/1910.13461](https://arxiv.org/abs/1910.13461)
+
+## Introduction
+
+BART is sequence-to-sequence model trained with denoising as pretraining objective. We show that this pretraining objective is more generic and show that we can match [RoBERTa](../roberta) results on SQuAD and GLUE and gain state-of-the-art results on summarization (XSum, CNN dataset), long form generative question answering (ELI5) and dialog response genration (ConvAI2). See the associated paper for more details.
+
+## Pre-trained models
+
+Model | Description | # params | Download
+---|---|---|---
+`bart.base` | BART model with 6 encoder and decoder layers | 140M | [bart.base.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/bart.base.tar.gz)
+`bart.large` | BART model with 12 encoder and decoder layers | 400M | [bart.large.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/bart.large.tar.gz)
+`bart.large.mnli` | `bart.large` finetuned on `MNLI` | 400M | [bart.large.mnli.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/bart.large.mnli.tar.gz)
+`bart.large.cnn` | `bart.large` finetuned on `CNN-DM` | 400M | [bart.large.cnn.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/bart.large.cnn.tar.gz)
+`bart.large.xsum` | `bart.large` finetuned on `Xsum` | 400M | [bart.large.xsum.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/bart.large.xsum.tar.gz)
+
+## Results
+
+**[GLUE (Wang et al., 2019)](https://gluebenchmark.com/)**
+_(dev set, single model, single-task finetuning)_
+
+Model | MNLI | QNLI | QQP | RTE | SST-2 | MRPC | CoLA | STS-B
+---|---|---|---|---|---|---|---|---
+`roberta.large` | 90.2 | 94.7 | 92.2 | 86.6 | 96.4 | 90.9 | 68.0 | 92.4
+`bart.large` | 89.9 | 94.9 | 92.5 | 87.0 | 96.6 | 90.4 | 62.8 | 91.2
+
+**[SQuAD (Rajpurkar et al., 2018)](https://rajpurkar.github.io/SQuAD-explorer/)**
+_(dev set, no additional data used)_
+
+Model | SQuAD 1.1 EM/F1 | SQuAD 2.0 EM/F1
+---|---|---
+`roberta.large` | 88.9/94.6 | 86.5/89.4
+`bart.large` | 88.8/94.6 | 86.1/89.2
+
+**[CNN/Daily Mail](http://nlpprogress.com/english/summarization.html)**
+_(test set, no additional data used)_
+
+Model | R1 | R2 | RL
+---|---|---|---
+`BERTSUMEXTABS` | 42.13 | 19.60 | 39.18
+`bart.large` | 44.16 | 21.28 | 40.90
+
+## Example usage
+
+##### Load BART from torch.hub (PyTorch >= 1.1):
+```python
+import torch
+bart = torch.hub.load('pytorch/fairseq', 'bart.large')
+bart.eval()  # disable dropout (or leave in train mode to finetune)
+```
+
+##### Load BART (for PyTorch 1.0 or custom models):
+```python
+# Download bart.large model
+wget https://dl.fbaipublicfiles.com/fairseq/models/bart.large.tar.gz
+tar -xzvf bart.large.tar.gz
+
+# Load the model in fairseq
+from fairseq.models.bart import BARTModel
+bart = BARTModel.from_pretrained('/path/to/bart.large', checkpoint_file='model.pt')
+bart.eval()  # disable dropout (or leave in train mode to finetune)
+```
+
+##### Apply Byte-Pair Encoding (BPE) to input text:
+```python
+tokens = bart.encode('Hello world!')
+assert tokens.tolist() == [0, 31414, 232, 328, 2]
+bart.decode(tokens)  # 'Hello world!'
+```
+
+##### Extract features from BART:
+```python
+# Extract the last layer's features
+last_layer_features = bart.extract_features(tokens)
+assert last_layer_features.size() == torch.Size([1, 5, 1024])
+
+# Extract all layer's features from decoder (layer 0 is the embedding layer)
+all_layers = bart.extract_features(tokens, return_all_hiddens=True)
+assert len(all_layers) == 13
+assert torch.all(all_layers[-1] == last_layer_features)
+```
+
+##### Use BART for sentence-pair classification tasks:
+```python
+# Download BART already finetuned for MNLI
+bart = torch.hub.load('pytorch/fairseq', 'bart.large.mnli')
+bart.eval()  # disable dropout for evaluation
+
+# Encode a pair of sentences and make a prediction
+tokens = bart.encode('BART is a seq2seq model.', 'BART is not sequence to sequence.')
+bart.predict('mnli', tokens).argmax()  # 0: contradiction
+
+# Encode another pair of sentences
+tokens = bart.encode('BART is denoising autoencoder.', 'BART is version of autoencoder.')
+bart.predict('mnli', tokens).argmax()  # 2: entailment
+```
+
+##### Register a new (randomly initialized) classification head:
+```python
+bart.register_classification_head('new_task', num_classes=3)
+logprobs = bart.predict('new_task', tokens)
+```
+
+##### Batched prediction:
+```python
+import torch
+from fairseq.data.data_utils import collate_tokens
+
+bart = torch.hub.load('pytorch/fairseq', 'bart.large.mnli')
+bart.eval()
+
+batch_of_pairs = [
+    ['BART is a seq2seq model.', 'BART is not sequence to sequence.'],
+    ['BART is denoising autoencoder.', 'BART is version of autoencoder.'],
+]
+
+batch = collate_tokens(
+    [bart.encode(pair[0], pair[1]) for pair in batch_of_pairs], pad_idx=1
+)
+
+logprobs = bart.predict('mnli', batch)
+print(logprobs.argmax(dim=1))
+# tensor([0, 2])
+```
+
+##### Using the GPU:
+```python
+bart.cuda()
+bart.predict('new_task', tokens)
+```
+
+#### Filling masks:
+
+BART can be used to fill multiple `<mask>` tokens in the input.
+```python
+bart = torch.hub.load('pytorch/fairseq', 'bart.base')
+bart.eval()
+bart.fill_mask(['The cat <mask> on the <mask>.'], topk=3, beam=10)
+# [[('The cat was on the ground.', tensor(-0.6183)), ('The cat was on the floor.', tensor(-0.6798)), ('The cat sleeps on the couch.', tensor(-0.6830))]]
+```
+
+Note that by default we enforce the output length to match the input length.
+This can be disabled by setting ``match_source_len=False``:
+```
+bart.fill_mask(['The cat <mask> on the <mask>.'], topk=3, beam=10, match_source_len=False)
+# [[('The cat was on the ground.', tensor(-0.6185)), ('The cat was asleep on the couch.', tensor(-0.6276)), ('The cat was on the floor.', tensor(-0.6800))]]
+```
+
+Example code to fill masks for a batch of sentences using GPU
+```
+bart.cuda()
+bart.fill_mask(['The cat <mask> on the <mask>.', 'The dog <mask> on the <mask>.'], topk=3, beam=10)
+# [[('The cat was on the ground.', tensor(-0.6183)), ('The cat was on the floor.', tensor(-0.6798)), ('The cat sleeps on the couch.', tensor(-0.6830))], [('The dog was on the ground.', tensor(-0.6190)), ('The dog lay on the ground.', tensor(-0.6711)),
+('The dog was asleep on the couch', tensor(-0.6796))]]
+```
+
+#### Evaluating the `bart.large.mnli` model:
+
+Example python code snippet to evaluate accuracy on the MNLI `dev_matched` set.
+```python
+label_map = {0: 'contradiction', 1: 'neutral', 2: 'entailment'}
+ncorrect, nsamples = 0, 0
+bart.cuda()
+bart.eval()
+with open('glue_data/MNLI/dev_matched.tsv') as fin:
+    fin.readline()
+    for index, line in enumerate(fin):
+        tokens = line.strip().split('\t')
+        sent1, sent2, target = tokens[8], tokens[9], tokens[-1]
+        tokens = bart.encode(sent1, sent2)
+        prediction = bart.predict('mnli', tokens).argmax().item()
+        prediction_label = label_map[prediction]
+        ncorrect += int(prediction_label == target)
+        nsamples += 1
+        print('| Accuracy: ', float(ncorrect)/float(nsamples))
+# Expected output: 0.9010
+```
+
+#### Evaluating the `bart.large.cnn` model:
+- Follow instructions [here](https://github.com/abisee/cnn-dailymail) to download and process into data-files such that `test.source` and `test.target` has one line for each non-tokenized sample.
+- For simpler preprocessing, you can also `wget https://cdn-datasets.huggingface.co/summarization/cnn_dm_v2.tgz`, although there is no guarantee of identical scores
+- `huggingface/transformers` has a simpler interface that supports [single-gpu](https://github.com/huggingface/transformers/blob/master/examples/legacy/seq2seq/run_eval.py) and [multi-gpu](https://github.com/huggingface/transformers/blob/master/examples/legacy/seq2seq/run_distributed_eval.py) beam search.
+    In `huggingface/transformers`, the BART models' paths are `facebook/bart-large-cnn` and `facebook/bart-large-xsum`.
+
+In `fairseq`, summaries can be generated using:
+
+```bash
+cp data-bin/cnn_dm/dict.source.txt  checkpoints/
+python examples/bart/summarize.py \
+  --model-dir pytorch/fairseq \
+  --model-file bart.large.cnn \
+  --src cnn_dm/test.source \
+  --out cnn_dm/test.hypo
+```
+
+For calculating rouge, install `files2rouge` from [here](https://github.com/pltrdy/files2rouge).
+
+```bash
+export CLASSPATH=/path/to/stanford-corenlp-full-2016-10-31/stanford-corenlp-3.7.0.jar
+
+# Tokenize hypothesis and target files.
+cat test.hypo | java edu.stanford.nlp.process.PTBTokenizer -ioFileList -preserveLines > test.hypo.tokenized
+cat test.target | java edu.stanford.nlp.process.PTBTokenizer -ioFileList -preserveLines > test.hypo.target
+files2rouge test.hypo.tokenized test.hypo.target
+# Expected output: (ROUGE-2 Average_F: 0.21238)
+```
+
+
+## Finetuning
+
+- [Finetuning on GLUE](README.glue.md)
+- [Finetuning on CNN-DM](README.summarization.md)
+
+## Citation
+
+```bibtex
+@article{lewis2019bart,
+    title = {BART: Denoising Sequence-to-Sequence Pre-training for Natural
+Language Generation, Translation, and Comprehension},
+    author = {Mike Lewis and Yinhan Liu and Naman Goyal and Marjan Ghazvininejad and
+              Abdelrahman Mohamed and Omer Levy and Veselin Stoyanov
+              and Luke Zettlemoyer },
+    journal={arXiv preprint arXiv:1910.13461},
+    year = {2019},
+}
+```

fairseq/examples/bart/README.summarization.md

@@ -0,0 +1,102 @@
+# Fine-tuning BART on CNN-Dailymail summarization task
+
+### 1) Download the CNN and Daily Mail data and preprocess it into data files with non-tokenized cased samples.
+
+Follow the instructions [here](https://github.com/abisee/cnn-dailymail) to download the original CNN and Daily Mail datasets. To preprocess the data, refer to the pointers in [this issue](https://github.com/pytorch/fairseq/issues/1391) or check out the code [here](https://github.com/artmatsak/cnn-dailymail).
+
+Follow the instructions [here](https://github.com/EdinburghNLP/XSum) to download the original Extreme Summarization datasets, or check out the code [here](https://github.com/EdinburghNLP/XSum/tree/master/XSum-Dataset), Please keep the raw dataset and make sure no tokenization nor BPE on the dataset.
+
+### 2) BPE preprocess:
+
+```bash
+wget -N 'https://dl.fbaipublicfiles.com/fairseq/gpt2_bpe/encoder.json'
+wget -N 'https://dl.fbaipublicfiles.com/fairseq/gpt2_bpe/vocab.bpe'
+wget -N 'https://dl.fbaipublicfiles.com/fairseq/gpt2_bpe/dict.txt'
+
+TASK=cnn_dm
+for SPLIT in train val
+do
+  for LANG in source target
+  do
+    python -m examples.roberta.multiprocessing_bpe_encoder \
+    --encoder-json encoder.json \
+    --vocab-bpe vocab.bpe \
+    --inputs "$TASK/$SPLIT.$LANG" \
+    --outputs "$TASK/$SPLIT.bpe.$LANG" \
+    --workers 60 \
+    --keep-empty;
+  done
+done
+```
+
+### 3) Binarize dataset:
+```bash
+fairseq-preprocess \
+  --source-lang "source" \
+  --target-lang "target" \
+  --trainpref "${TASK}/train.bpe" \
+  --validpref "${TASK}/val.bpe" \
+  --destdir "${TASK}-bin/" \
+  --workers 60 \
+  --srcdict dict.txt \
+  --tgtdict dict.txt;
+```
+
+### 4) Fine-tuning on CNN-DM summarization task:
+Example fine-tuning CNN-DM
+```bash
+TOTAL_NUM_UPDATES=20000  
+WARMUP_UPDATES=500      
+LR=3e-05
+MAX_TOKENS=2048
+UPDATE_FREQ=4
+BART_PATH=/path/to/bart/model.pt
+
+CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 fairseq-train cnn_dm-bin \
+    --restore-file $BART_PATH \
+    --max-tokens $MAX_TOKENS \
+    --task translation \
+    --source-lang source --target-lang target \
+    --truncate-source \
+    --layernorm-embedding \
+    --share-all-embeddings \
+    --share-decoder-input-output-embed \
+    --reset-optimizer --reset-dataloader --reset-meters \
+    --required-batch-size-multiple 1 \
+    --arch bart_large \
+    --criterion label_smoothed_cross_entropy \
+    --label-smoothing 0.1 \
+    --dropout 0.1 --attention-dropout 0.1 \
+    --weight-decay 0.01 --optimizer adam --adam-betas "(0.9, 0.999)" --adam-eps 1e-08 \
+    --clip-norm 0.1 \
+    --lr-scheduler polynomial_decay --lr $LR --total-num-update $TOTAL_NUM_UPDATES --warmup-updates $WARMUP_UPDATES \
+    --fp16 --update-freq $UPDATE_FREQ \
+    --skip-invalid-size-inputs-valid-test \
+    --find-unused-parameters;
+```
+Above is expected to run on `1` node with `8 32gb-V100`.
+Expected training time is about `5 hours`. Training time can be reduced with distributed training on `4` nodes and `--update-freq 1`.
+
+Use TOTAL_NUM_UPDATES=15000 UPDATE_FREQ=2 for Xsum task
+
+### Inference for CNN-DM test data using above trained checkpoint.
+After training the model as mentioned in previous step, you can perform inference with checkpoints in `checkpoints/` directory using `eval_cnn.py`, for example
+
+```bash
+cp data-bin/cnn_dm/dict.source.txt  checkpoints/
+python examples/bart/summarize.py \
+  --model-dir checkpoints \
+  --model-file checkpoint_best.pt \
+  --src cnn_dm/test.source \
+  --out cnn_dm/test.hypo
+```
+For XSUM, which uses beam=6, lenpen=1.0, max_len_b=60, min_len=10:
+```bash
+cp data-bin/cnn_dm/dict.source.txt  checkpoints/
+python examples/bart/summarize.py \
+  --model-dir checkpoints \
+  --model-file checkpoint_best.pt \
+  --src cnn_dm/test.source \
+  --out cnn_dm/test.hypo \
+  --xsum-kwargs
+```

fairseq/examples/bart/summarize.py

@@ -0,0 +1,100 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+from fairseq.models.bart import BARTModel
+import argparse
+
+XSUM_KWARGS = dict(beam=6, lenpen=1.0, max_len_b=60, min_len=10, no_repeat_ngram_size=3)
+CNN_KWARGS = dict(beam=4, lenpen=2.0, max_len_b=140, min_len=55, no_repeat_ngram_size=3)
+
+
+@torch.no_grad()
+def generate(bart, infile, outfile="bart_hypo.txt", bsz=32, n_obs=None, **eval_kwargs):
+    count = 1
+
+    # if n_obs is not None: bsz = min(bsz, n_obs)
+
+    with open(infile) as source, open(outfile, "w") as fout:
+        sline = source.readline().strip()
+        slines = [sline]
+        for sline in source:
+            if n_obs is not None and count > n_obs:
+                break
+            if count % bsz == 0:
+                hypotheses_batch = bart.sample(slines, **eval_kwargs)
+                for hypothesis in hypotheses_batch:
+                    fout.write(hypothesis + "\n")
+                    fout.flush()
+                slines = []
+
+            slines.append(sline.strip())
+            count += 1
+
+        if slines != []:
+            hypotheses_batch = bart.sample(slines, **eval_kwargs)
+            for hypothesis in hypotheses_batch:
+                fout.write(hypothesis + "\n")
+                fout.flush()
+
+
+def main():
+    """
+    Usage::
+
+         python examples/bart/summarize.py \
+            --model-dir $HOME/bart.large.cnn \
+            --model-file model.pt \
+            --src $HOME/data-bin/cnn_dm/test.source
+    """
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model-dir",
+        required=True,
+        type=str,
+        default="bart.large.cnn/",
+        help="path containing model file and src_dict.txt",
+    )
+    parser.add_argument(
+        "--model-file",
+        default="checkpoint_best.pt",
+        help="where in model_dir are weights saved",
+    )
+    parser.add_argument(
+        "--src", default="test.source", help="text to summarize", type=str
+    )
+    parser.add_argument(
+        "--out", default="test.hypo", help="where to save summaries", type=str
+    )
+    parser.add_argument("--bsz", default=32, help="where to save summaries", type=int)
+    parser.add_argument(
+        "--n", default=None, help="how many examples to summarize", type=int
+    )
+    parser.add_argument(
+        "--xsum-kwargs",
+        action="store_true",
+        default=False,
+        help="if true use XSUM_KWARGS else CNN_KWARGS",
+    )
+    args = parser.parse_args()
+    eval_kwargs = XSUM_KWARGS if args.xsum_kwargs else CNN_KWARGS
+    if args.model_dir == "pytorch/fairseq":
+        bart = torch.hub.load("pytorch/fairseq", args.model_file)
+    else:
+        bart = BARTModel.from_pretrained(
+            args.model_dir,
+            checkpoint_file=args.model_file,
+            data_name_or_path=args.model_dir,
+        )
+    bart = bart.eval()
+    if torch.cuda.is_available():
+        bart = bart.cuda().half()
+    generate(
+        bart, args.src, bsz=args.bsz, n_obs=args.n, outfile=args.out, **eval_kwargs
+    )
+
+
+if __name__ == "__main__":
+    main()

fairseq/examples/byte_level_bpe/README.md

@@ -0,0 +1,88 @@
+# Neural Machine Translation with Byte-Level Subwords
+
+https://arxiv.org/abs/1909.03341
+
+We provide an implementation of byte-level byte-pair encoding (BBPE), taking IWSLT 2017 Fr-En translation as
+example.
+
+## Data
+Get data and generate fairseq binary dataset:
+```bash
+bash ./get_data.sh
+```
+
+## Model Training
+Train Transformer model with Bi-GRU embedding contextualization (implemented in `gru_transformer.py`):
+```bash
+# VOCAB=bytes
+# VOCAB=chars
+VOCAB=bbpe2048
+# VOCAB=bpe2048
+# VOCAB=bbpe4096
+# VOCAB=bpe4096
+# VOCAB=bpe16384
+```
+```bash
+fairseq-train "data/bin_${VOCAB}" --task translation --user-dir examples/byte_level_bpe/gru_transformer \
+    --arch gru_transformer --encoder-layers 2 --decoder-layers 2 --dropout 0.3 --share-all-embeddings \
+    --optimizer adam --adam-betas '(0.9, 0.98)' \
+    --lr 5e-4 --lr-scheduler inverse_sqrt --warmup-updates 4000 \
+    --criterion label_smoothed_cross_entropy --label-smoothing 0.1 \
+    --log-format 'simple' --log-interval 100 --save-dir "checkpoints/${VOCAB}" \
+    --batch-size 100 --max-update 100000 --update-freq 2
+```
+
+## Generation
+`fairseq-generate` requires bytes (BBPE) decoder to convert byte-level representation back to characters:
+```bash
+# BPE=--bpe bytes
+# BPE=--bpe characters
+BPE=--bpe byte_bpe --sentencepiece-model-path data/spm_bbpe2048.model
+# BPE=--bpe sentencepiece --sentencepiece-model data/spm_bpe2048.model
+# BPE=--bpe byte_bpe --sentencepiece-model-path data/spm_bbpe4096.model
+# BPE=--bpe sentencepiece --sentencepiece-model data/spm_bpe4096.model
+# BPE=--bpe sentencepiece --sentencepiece-model data/spm_bpe16384.model
+```
+
+```bash
+fairseq-generate "data/bin_${VOCAB}" --task translation --user-dir examples/byte_level_bpe/gru_transformer \
+    --source-lang fr --gen-subset test --sacrebleu --path "checkpoints/${VOCAB}/checkpoint_last.pt" \
+    --tokenizer moses --moses-target-lang en ${BPE}
+```
+When using `fairseq-interactive`, bytes (BBPE) encoder/decoder is required to tokenize input data and detokenize model predictions:
+```bash
+fairseq-interactive "data/bin_${VOCAB}" --task translation --user-dir examples/byte_level_bpe/gru_transformer \
+    --path "checkpoints/${VOCAB}/checkpoint_last.pt" --input data/test.fr --tokenizer moses --moses-source-lang fr \
+    --moses-target-lang en ${BPE} --buffer-size 1000 --max-tokens 10000
+```
+
+## Results
+| Vocabulary    | Model  | BLEU |
+|:-------------:|:-------------:|:-------------:|
+| Joint BPE 16k ([Kudo, 2018](https://arxiv.org/abs/1804.10959)) | 512d LSTM 2+2 | 33.81 |
+| Joint BPE 16k | Transformer base 2+2 (w/ GRU) | 36.64 (36.72) |
+| Joint BPE 4k | Transformer base 2+2 (w/ GRU) | 35.49 (36.10) |
+| Joint BBPE 4k | Transformer base 2+2 (w/ GRU) | 35.61 (35.82) |
+| Joint BPE 2k | Transformer base 2+2 (w/ GRU) | 34.87 (36.13) |
+| Joint BBPE 2k | Transformer base 2+2 (w/ GRU) | 34.98 (35.43) |
+| Characters | Transformer base 2+2 (w/ GRU) | 31.78 (33.30) |
+| Bytes | Transformer base 2+2 (w/ GRU) | 31.57 (33.62) |
+
+
+## Citation
+```
+@misc{wang2019neural,
+    title={Neural Machine Translation with Byte-Level Subwords},
+    author={Changhan Wang and Kyunghyun Cho and Jiatao Gu},
+    year={2019},
+    eprint={1909.03341},
+    archivePrefix={arXiv},
+    primaryClass={cs.CL}
+}
+```
+
+
+## Contact
+Changhan Wang ([[email protected]](mailto:[email protected])),
+Kyunghyun Cho ([[email protected]](mailto:[email protected])),
+Jiatao Gu ([[email protected]](mailto:[email protected]))

fairseq/examples/byte_level_bpe/get_bitext.py

@@ -0,0 +1,254 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+
+import argparse
+import os
+import os.path as op
+from collections import namedtuple
+from multiprocessing import cpu_count
+from typing import List, Optional
+
+import sentencepiece as sp
+from fairseq.data.encoders.byte_bpe import ByteBPE
+from fairseq.data.encoders.byte_utils import byte_encode
+from fairseq.data.encoders.bytes import Bytes
+from fairseq.data.encoders.characters import Characters
+from fairseq.data.encoders.moses_tokenizer import MosesTokenizer
+from fairseq.data.encoders.sentencepiece_bpe import SentencepieceBPE
+
+
+SPLITS = ["train", "valid", "test"]
+
+
+def _convert_xml(in_path: str, out_path: str):
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            ss = s.strip()
+            if not ss.startswith("<seg"):
+                continue
+            ss = ss.replace("</seg>", "").split('">')
+            assert len(ss) == 2
+            f_o.write(ss[1].strip() + "\n")
+
+
+def _convert_train(in_path: str, out_path: str):
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            ss = s.strip()
+            if ss.startswith("<"):
+                continue
+            f_o.write(ss.strip() + "\n")
+
+
+def _get_bytes(in_path: str, out_path: str):
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            f_o.write(Bytes.encode(s.strip()) + "\n")
+
+
+def _get_chars(in_path: str, out_path: str):
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            f_o.write(Characters.encode(s.strip()) + "\n")
+
+
+def pretokenize(in_path: str, out_path: str, src: str, tgt: str):
+    Args = namedtuple(
+        "Args",
+        [
+            "moses_source_lang",
+            "moses_target_lang",
+            "moses_no_dash_splits",
+            "moses_no_escape",
+        ],
+    )
+    args = Args(
+        moses_source_lang=src,
+        moses_target_lang=tgt,
+        moses_no_dash_splits=False,
+        moses_no_escape=False,
+    )
+    pretokenizer = MosesTokenizer(args)
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            f_o.write(pretokenizer.encode(s.strip()) + "\n")
+
+
+def _convert_to_bchar(in_path_prefix: str, src: str, tgt: str, out_path: str):
+    with open(out_path, "w") as f_o:
+        for lang in [src, tgt]:
+            with open(f"{in_path_prefix}.{lang}") as f:
+                for s in f:
+                    f_o.write(byte_encode(s.strip()) + "\n")
+
+
+def _get_bpe(in_path: str, model_prefix: str, vocab_size: int):
+    arguments = [
+        f"--input={in_path}",
+        f"--model_prefix={model_prefix}",
+        f"--model_type=bpe",
+        f"--vocab_size={vocab_size}",
+        "--character_coverage=1.0",
+        "--normalization_rule_name=identity",
+        f"--num_threads={cpu_count()}",
+    ]
+    sp.SentencePieceTrainer.Train(" ".join(arguments))
+
+
+def _apply_bbpe(model_path: str, in_path: str, out_path: str):
+    Args = namedtuple("Args", ["sentencepiece_model_path"])
+    args = Args(sentencepiece_model_path=model_path)
+    tokenizer = ByteBPE(args)
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            f_o.write(tokenizer.encode(s.strip()) + "\n")
+
+
+def _apply_bpe(model_path: str, in_path: str, out_path: str):
+    Args = namedtuple("Args", ["sentencepiece_model"])
+    args = Args(sentencepiece_model=model_path)
+    tokenizer = SentencepieceBPE(args)
+    with open(in_path) as f, open(out_path, "w") as f_o:
+        for s in f:
+            f_o.write(tokenizer.encode(s.strip()) + "\n")
+
+
+def _concat_files(in_paths: List[str], out_path: str):
+    with open(out_path, "w") as f_o:
+        for p in in_paths:
+            with open(p) as f:
+                for r in f:
+                    f_o.write(r)
+
+
+def preprocess_iwslt17(
+    root: str,
+    src: str,
+    tgt: str,
+    bpe_size: Optional[int],
+    need_chars: bool,
+    bbpe_size: Optional[int],
+    need_bytes: bool,
+):
+    # extract bitext
+    in_root = op.join(root, f"{src}-{tgt}")
+    for lang in [src, tgt]:
+        _convert_train(
+            op.join(in_root, f"train.tags.{src}-{tgt}.{lang}"),
+            op.join(root, f"train.{lang}"),
+        )
+        _convert_xml(
+            op.join(in_root, f"IWSLT17.TED.dev2010.{src}-{tgt}.{lang}.xml"),
+            op.join(root, f"valid.{lang}"),
+        )
+        _convert_xml(
+            op.join(in_root, f"IWSLT17.TED.tst2015.{src}-{tgt}.{lang}.xml"),
+            op.join(root, f"test.{lang}"),
+        )
+    # pre-tokenize
+    for lang in [src, tgt]:
+        for split in SPLITS:
+            pretokenize(
+                op.join(root, f"{split}.{lang}"),
+                op.join(root, f"{split}.moses.{lang}"),
+                src,
+                tgt,
+            )
+    # tokenize with BPE vocabulary
+    if bpe_size is not None:
+        # learn vocabulary
+        concated_train_path = op.join(root, "train.all")
+        _concat_files(
+            [op.join(root, "train.moses.fr"), op.join(root, "train.moses.en")],
+            concated_train_path,
+        )
+        bpe_model_prefix = op.join(root, f"spm_bpe{bpe_size}")
+        _get_bpe(concated_train_path, bpe_model_prefix, bpe_size)
+        os.remove(concated_train_path)
+        # apply
+        for lang in [src, tgt]:
+            for split in SPLITS:
+                _apply_bpe(
+                    bpe_model_prefix + ".model",
+                    op.join(root, f"{split}.moses.{lang}"),
+                    op.join(root, f"{split}.moses.bpe{bpe_size}.{lang}"),
+                )
+    # tokenize with bytes vocabulary
+    if need_bytes:
+        for lang in [src, tgt]:
+            for split in SPLITS:
+                _get_bytes(
+                    op.join(root, f"{split}.moses.{lang}"),
+                    op.join(root, f"{split}.moses.bytes.{lang}"),
+                )
+    # tokenize with characters vocabulary
+    if need_chars:
+        for lang in [src, tgt]:
+            for split in SPLITS:
+                _get_chars(
+                    op.join(root, f"{split}.moses.{lang}"),
+                    op.join(root, f"{split}.moses.chars.{lang}"),
+                )
+    # tokenize with byte-level BPE vocabulary
+    if bbpe_size is not None:
+        # learn vocabulary
+        bchar_path = op.join(root, "train.bchar")
+        _convert_to_bchar(op.join(root, "train.moses"), src, tgt, bchar_path)
+        bbpe_model_prefix = op.join(root, f"spm_bbpe{bbpe_size}")
+        _get_bpe(bchar_path, bbpe_model_prefix, bbpe_size)
+        os.remove(bchar_path)
+        # apply
+        for lang in [src, tgt]:
+            for split in SPLITS:
+                _apply_bbpe(
+                    bbpe_model_prefix + ".model",
+                    op.join(root, f"{split}.moses.{lang}"),
+                    op.join(root, f"{split}.moses.bbpe{bbpe_size}.{lang}"),
+                )
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--root", type=str, default="data")
+    parser.add_argument(
+        "--bpe-vocab",
+        default=None,
+        type=int,
+        help="Generate tokenized bitext with BPE of size K."
+        "Default to None (disabled).",
+    )
+    parser.add_argument(
+        "--bbpe-vocab",
+        default=None,
+        type=int,
+        help="Generate tokenized bitext with BBPE of size K."
+        "Default to None (disabled).",
+    )
+    parser.add_argument(
+        "--byte-vocab",
+        action="store_true",
+        help="Generate tokenized bitext with bytes vocabulary",
+    )
+    parser.add_argument(
+        "--char-vocab",
+        action="store_true",
+        help="Generate tokenized bitext with chars vocabulary",
+    )
+    args = parser.parse_args()
+
+    preprocess_iwslt17(
+        args.root,
+        "fr",
+        "en",
+        args.bpe_vocab,
+        args.char_vocab,
+        args.bbpe_vocab,
+        args.byte_vocab,
+    )
+
+
+if __name__ == "__main__":
+    main()

fairseq/examples/byte_level_bpe/get_data.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+PY_BIN_ROOT=
+
+# PyPI dependency
+${PY_BIN_ROOT}pip install sentencepiece sacremoses
+
+# Get data
+if [ ! -d "data" ]; then
+  mkdir data
+fi
+
+if [ ! -f "data/fr-en.tgz" ]; then
+  wget https://wit3.fbk.eu/archive/2017-01-trnted/texts/fr/en/fr-en.tgz -P data
+  tar xvf data/fr-en.tgz -C data
+fi
+${PY_BIN_ROOT}python get_bitext.py --bpe-vocab 16384 --byte-vocab --char-vocab
+for VOCAB_SIZE in 2048 4096; do
+  ${PY_BIN_ROOT}python get_bitext.py --bpe-vocab ${VOCAB_SIZE} --bbpe-vocab ${VOCAB_SIZE}
+done
+rm -r data/fr-en data/fr-en.tgz
+
+# Generate binary dataset
+${PY_BIN_ROOT}/fairseq-preprocess --source-lang fr --target-lang en --destdir data/bin_bpe16384 --joined-dictionary \
+  --workers "$(nproc)" --trainpref data/train.moses.bpe16384 --validpref data/valid.moses.bpe16384 \
+  --testpref data/test.moses.bpe16384
+
+${PY_BIN_ROOT}/fairseq-preprocess --source-lang fr --target-lang en --destdir data/bin_bytes --joined-dictionary \
+  --workers "$(nproc)" --trainpref data/train.moses.bytes --validpref data/valid.moses.bytes \
+  --testpref data/test.moses.bytes
+
+${PY_BIN_ROOT}/fairseq-preprocess --source-lang fr --target-lang en --destdir data/bin_chars --joined-dictionary \
+  --workers "$(nproc)" --trainpref data/train.moses.chars --validpref data/valid.moses.chars \
+  --testpref data/test.moses.chars
+
+for VOCAB_SIZE in 2048 4096; do
+  for TYPE in bbpe bpe; do
+    ${PY_BIN_ROOT}/fairseq-preprocess --source-lang fr --target-lang en --destdir "data/bin_${TYPE}${VOCAB_SIZE}" \
+      --joined-dictionary --workers "$(nproc)" --trainpref "data/train.moses.${TYPE}${VOCAB_SIZE}" \
+      --validpref "data/valid.moses.${TYPE}${VOCAB_SIZE}" --testpref "data/test.moses.${TYPE}${VOCAB_SIZE}"
+  done
+done

fairseq/examples/byte_level_bpe/gru_transformer.py

@@ -0,0 +1,107 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch.nn as nn
+import torch.nn.functional as F
+from fairseq.models import register_model, register_model_architecture
+from fairseq.models.transformer import TransformerEncoder, TransformerModel
+
+
+@register_model("gru_transformer")
+class GRUTransformerModel(TransformerModel):
+    @classmethod
+    def build_encoder(cls, args, src_dict, embed_tokens):
+        return GRUTransformerEncoder(args, src_dict, embed_tokens)
+
+
+class GRUTransformerEncoder(TransformerEncoder):
+    def __init__(self, args, dictionary, embed_tokens):
+        super().__init__(args, dictionary, embed_tokens)
+        self.emb_ctx = nn.GRU(
+            input_size=embed_tokens.embedding_dim,
+            hidden_size=embed_tokens.embedding_dim // 2,
+            num_layers=1,
+            bidirectional=True,
+        )
+
+    def forward_embedding(self, src_tokens):
+        # embed tokens and positions
+        x = embed = self.embed_scale * self.embed_tokens(src_tokens)
+        if self.embed_positions is not None:
+            x = embed + self.embed_positions(src_tokens)
+
+        # contextualize embeddings
+        x = x.transpose(0, 1)
+        x = self.dropout_module(x)
+        x, _ = self.emb_ctx.forward(x)
+        x = x.transpose(0, 1)
+
+        if self.layernorm_embedding is not None:
+            x = self.layernorm_embedding(x)
+        x = self.dropout_module(x)
+        return x, embed
+
+
+@register_model_architecture("gru_transformer", "gru_transformer")
+def gru_transformer_base_architecture(args):
+    args.encoder_embed_path = getattr(args, "encoder_embed_path", None)
+    args.encoder_embed_dim = getattr(args, "encoder_embed_dim", 512)
+    args.encoder_ffn_embed_dim = getattr(args, "encoder_ffn_embed_dim", 2048)
+    args.encoder_layers = getattr(args, "encoder_layers", 6)
+    args.encoder_attention_heads = getattr(args, "encoder_attention_heads", 8)
+    args.encoder_normalize_before = getattr(args, "encoder_normalize_before", False)
+    args.encoder_learned_pos = getattr(args, "encoder_learned_pos", False)
+    args.decoder_embed_path = getattr(args, "decoder_embed_path", None)
+    args.decoder_embed_dim = getattr(args, "decoder_embed_dim", args.encoder_embed_dim)
+    args.decoder_ffn_embed_dim = getattr(
+        args, "decoder_ffn_embed_dim", args.encoder_ffn_embed_dim
+    )
+    args.decoder_layers = getattr(args, "decoder_layers", 6)
+    args.decoder_attention_heads = getattr(args, "decoder_attention_heads", 8)
+    args.decoder_normalize_before = getattr(args, "decoder_normalize_before", False)
+    args.decoder_learned_pos = getattr(args, "decoder_learned_pos", False)
+    args.attention_dropout = getattr(args, "attention_dropout", 0.0)
+    args.activation_dropout = getattr(args, "activation_dropout", 0.0)
+    args.activation_fn = getattr(args, "activation_fn", "relu")
+    args.dropout = getattr(args, "dropout", 0.1)
+    args.adaptive_softmax_cutoff = getattr(args, "adaptive_softmax_cutoff", None)
+    args.adaptive_softmax_dropout = getattr(args, "adaptive_softmax_dropout", 0)
+    args.share_decoder_input_output_embed = getattr(
+        args, "share_decoder_input_output_embed", False
+    )
+    args.share_all_embeddings = getattr(args, "share_all_embeddings", False)
+    args.no_token_positional_embeddings = getattr(
+        args, "no_token_positional_embeddings", False
+    )
+    args.adaptive_input = getattr(args, "adaptive_input", False)
+    args.no_cross_attention = getattr(args, "no_cross_attention", False)
+    args.cross_self_attention = getattr(args, "cross_self_attention", False)
+    args.layer_wise_attention = getattr(args, "layer_wise_attention", False)
+
+    args.decoder_output_dim = getattr(
+        args, "decoder_output_dim", args.decoder_embed_dim
+    )
+    args.decoder_input_dim = getattr(args, "decoder_input_dim", args.decoder_embed_dim)
+
+    args.no_scale_embedding = getattr(args, "no_scale_embedding", False)
+    args.layernorm_embedding = getattr(args, "layernorm_embedding", False)
+
+
+@register_model_architecture("gru_transformer", "gru_transformer_big")
+def gru_transformer_big(args):
+    args.encoder_embed_dim = getattr(args, "encoder_embed_dim", 1024)
+    args.encoder_ffn_embed_dim = getattr(args, "encoder_ffn_embed_dim", 4096)
+    args.encoder_attention_heads = getattr(args, "encoder_attention_heads", 16)
+    args.encoder_normalize_before = getattr(args, "encoder_normalize_before", False)
+    args.decoder_embed_dim = getattr(args, "decoder_embed_dim", 1024)
+    args.decoder_ffn_embed_dim = getattr(args, "decoder_ffn_embed_dim", 4096)
+    args.decoder_attention_heads = getattr(args, "decoder_attention_heads", 16)
+    args.dropout = getattr(args, "dropout", 0.3)
+    gru_transformer_base_architecture(args)

fairseq/examples/camembert/README.md

@@ -0,0 +1,75 @@
+# CamemBERT: a Tasty French Language Model
+
+## Introduction
+
+[CamemBERT](https://arxiv.org/abs/1911.03894) is a pretrained language model trained on 138GB of French text based on RoBERTa.
+
+Also available in [github.com/huggingface/transformers](https://github.com/huggingface/transformers/).
+
+## Pre-trained models
+
+| Model                          | #params | Download                                                                                                                 | Arch. | Training data                     |
+|--------------------------------|---------|--------------------------------------------------------------------------------------------------------------------------|-------|-----------------------------------|
+| `camembert` / `camembert-base` | 110M    | [camembert-base.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-base.tar.gz)                             | Base  | OSCAR (138 GB of text)            |
+| `camembert-large`              | 335M    | [camembert-large.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-large.tar.gz)                           | Large | CCNet (135 GB of text)            |
+| `camembert-base-ccnet`         | 110M    | [camembert-base-ccnet.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-base-ccnet.tar.gz)                 | Base  | CCNet (135 GB of text)            |
+| `camembert-base-wikipedia-4gb` | 110M    | [camembert-base-wikipedia-4gb.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-base-wikipedia-4gb.tar.gz) | Base  | Wikipedia (4 GB of text)          |
+| `camembert-base-oscar-4gb`     | 110M    | [camembert-base-oscar-4gb.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-base-oscar-4gb.tar.gz)         | Base  | Subsample of OSCAR (4 GB of text) |
+| `camembert-base-ccnet-4gb`     | 110M    | [camembert-base-ccnet-4gb.tar.gz](https://dl.fbaipublicfiles.com/fairseq/models/camembert-base-ccnet-4gb.tar.gz)         | Base  | Subsample of CCNet (4 GB of text) |
+
+## Example usage
+
+### fairseq
+##### Load CamemBERT from torch.hub (PyTorch >= 1.1):
+```python
+import torch
+camembert = torch.hub.load('pytorch/fairseq', 'camembert')
+camembert.eval()  # disable dropout (or leave in train mode to finetune)
+```
+
+##### Load CamemBERT (for PyTorch 1.0 or custom models):
+```python
+# Download camembert model
+wget https://dl.fbaipublicfiles.com/fairseq/models/camembert-base.tar.gz
+tar -xzvf camembert.tar.gz
+
+# Load the model in fairseq
+from fairseq.models.roberta import CamembertModel
+camembert = CamembertModel.from_pretrained('/path/to/camembert')
+camembert.eval()  # disable dropout (or leave in train mode to finetune)
+```
+
+##### Filling masks:
+```python
+masked_line = 'Le camembert est <mask> :)'
+camembert.fill_mask(masked_line, topk=3)
+# [('Le camembert est délicieux :)', 0.4909118115901947, ' délicieux'),
+#  ('Le camembert est excellent :)', 0.10556942224502563, ' excellent'),
+#  ('Le camembert est succulent :)', 0.03453322499990463, ' succulent')]
+```
+
+##### Extract features from Camembert:
+```python
+# Extract the last layer's features
+line = "J'aime le camembert !"
+tokens = camembert.encode(line)
+last_layer_features = camembert.extract_features(tokens)
+assert last_layer_features.size() == torch.Size([1, 10, 768])
+
+# Extract all layer's features (layer 0 is the embedding layer)
+all_layers = camembert.extract_features(tokens, return_all_hiddens=True)
+assert len(all_layers) == 13
+assert torch.all(all_layers[-1] == last_layer_features)
+```
+
+## Citation
+If you use our work, please cite:
+
+```bibtex
+@inproceedings{martin2020camembert,
+  title={CamemBERT: a Tasty French Language Model},
+  author={Martin, Louis and Muller, Benjamin and Su{\'a}rez, Pedro Javier Ortiz and Dupont, Yoann and Romary, Laurent and de la Clergerie, {\'E}ric Villemonte and Seddah, Djam{\'e} and Sagot, Beno{\^\i}t},
+  booktitle={Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
+  year={2020}
+}
+```

fairseq/examples/constrained_decoding/README.md

@@ -0,0 +1,123 @@
+# (Vectorized) Lexically constrained decoding with dynamic beam allocation
+
+This page provides instructions for how to use lexically constrained decoding in Fairseq.
+Fairseq implements the code described in the following papers:
+
+* [Fast Lexically Constrained Decoding With Dynamic Beam Allocation](https://www.aclweb.org/anthology/N18-1119/) (Post & Vilar, 2018)
+* [Improved Lexically Constrained Decoding for Translation and Monolingual Rewriting](https://www.aclweb.org/anthology/N19-1090/) (Hu et al., 2019)
+
+## Quick start
+
+Constrained search is enabled by adding the command-line argument `--constraints` to `fairseq-interactive`.
+Constraints are appended to each line of input, separated by tabs. Each constraint (one or more tokens)
+is a separate field.
+
+The following command, using [Fairseq's WMT19 German--English model](https://github.com/pytorch/fairseq/blob/main/examples/wmt19/README.md),
+translates the sentence *Die maschinelle Übersetzung ist schwer zu kontrollieren.* with the constraints
+"hard" and "to influence".
+
+    echo -e "Die maschinelle Übersetzung ist schwer zu kontrollieren.\thard\ttoinfluence" \
+    | normalize.py | tok.py \
+    | fairseq-interactive /path/to/model \
+      --path /path/to/model/model1.pt \
+      --bpe fastbpe \
+      --bpe-codes /path/to/model/bpecodes \
+      --constraints \
+      -s de -t en \
+      --beam 10
+
+(tok.py and normalize.py can be found in the same directory as this README; they are just shortcuts around Fairseq's WMT19 preprocessing).
+This will generate the following output:
+
+    [snip]
+    S-0     Die masch@@ in@@ elle Über@@ setzung ist schwer zu kontrollieren .
+    W-0     1.844   seconds
+    C-0     hard
+    C-0     influence
+    H-0     -1.5333266258239746     Mach@@ ine trans@@ lation is hard to influence .
+    D-0     -1.5333266258239746     Machine translation is hard to influence .
+    P-0     -0.5434 -0.1423 -0.1930 -0.1415 -0.2346 -1.8031 -0.1701 -11.7727 -0.1815 -0.1511
+
+By default, constraints are generated in the order supplied, with any number (zero or more) of tokens generated
+between constraints. If you wish for the decoder to order the constraints, then use `--constraints unordered`.
+Note that you may want to use a larger beam.
+
+## Implementation details
+
+The heart of the implementation is in `fairseq/search.py`, which adds a `LexicallyConstrainedBeamSearch` instance.
+This instance of beam search tracks the progress of each hypothesis in the beam through the set of constraints
+provided for each input sentence. It does this using one of two classes, both found in `fairseq/token_generation_contstraints.py`:
+
+* OrderedConstraintState: assumes the `C` input constraints will be generated in the provided order
+* UnorderedConstraintState: tries to apply `C` (phrasal) constraints in all `C!` orders
+
+## Differences from Sockeye
+
+There are a number of [differences from Sockeye's implementation](https://awslabs.github.io/sockeye/inference.html#lexical-constraints).
+
+* Generating constraints in the order supplied (the default option here) is not available in Sockeye.
+* Due to an improved beam allocation method, there is no need to prune the beam.
+* Again due to better allocation, beam sizes as low as 10 or even 5 are often sufficient.
+* [The vector extensions described in Hu et al.](https://github.com/edwardjhu/sockeye/tree/trie_constraints) (NAACL 2019) were never merged
+  into the main Sockeye branch.
+
+## Citation
+
+The paper first describing lexical constraints for seq2seq decoding is:
+
+```bibtex
+@inproceedings{hokamp-liu-2017-lexically,
+  title = "Lexically Constrained Decoding for Sequence Generation Using Grid Beam Search",
+  author = "Hokamp, Chris  and
+    Liu, Qun",
+  booktitle = "Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
+  month = jul,
+  year = "2017",
+  address = "Vancouver, Canada",
+  publisher = "Association for Computational Linguistics",
+  url = "https://www.aclweb.org/anthology/P17-1141",
+  doi = "10.18653/v1/P17-1141",
+  pages = "1535--1546",
+}
+```
+
+The fairseq implementation uses the extensions described in
+
+```bibtex
+@inproceedings{post-vilar-2018-fast,
+    title = "Fast Lexically Constrained Decoding with Dynamic Beam Allocation for Neural Machine Translation",
+    author = "Post, Matt  and
+      Vilar, David",
+    booktitle = "Proceedings of the 2018 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers)",
+    month = jun,
+    year = "2018",
+    address = "New Orleans, Louisiana",
+    publisher = "Association for Computational Linguistics",
+    url = "https://www.aclweb.org/anthology/N18-1119",
+    doi = "10.18653/v1/N18-1119",
+    pages = "1314--1324",
+}
+```
+
+and
+
+```bibtex
+@inproceedings{hu-etal-2019-improved,
+  title = "Improved Lexically Constrained Decoding for Translation and Monolingual Rewriting",
+  author = "Hu, J. Edward  and
+    Khayrallah, Huda  and
+    Culkin, Ryan  and
+    Xia, Patrick  and
+    Chen, Tongfei  and
+    Post, Matt  and
+    Van Durme, Benjamin",
+  booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
+  month = jun,
+  year = "2019",
+  address = "Minneapolis, Minnesota",
+  publisher = "Association for Computational Linguistics",
+  url = "https://www.aclweb.org/anthology/N19-1090",
+  doi = "10.18653/v1/N19-1090",
+  pages = "839--850",
+}
+```

fairseq/examples/constrained_decoding/normalize.py

@@ -0,0 +1,27 @@
+#!/usr/bin/env python3
+#
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import sys
+
+from sacremoses.normalize import MosesPunctNormalizer
+
+
+def main(args):
+    normalizer = MosesPunctNormalizer(lang=args.lang, penn=args.penn)
+    for line in sys.stdin:
+        print(normalizer.normalize(line.rstrip()), flush=True)
+
+
+if __name__ == "__main__":
+    import argparse
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--lang", "-l", default="en")
+    parser.add_argument("--penn", "-p", action="store_true")
+    args = parser.parse_args()
+
+    main(args)

fairseq/examples/constrained_decoding/tok.py

@@ -0,0 +1,34 @@
+#!/usr/bin/env python3
+#
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+
+import sys
+
+import sacremoses
+
+
+def main(args):
+    """Tokenizes, preserving tabs"""
+    mt = sacremoses.MosesTokenizer(lang=args.lang)
+
+    def tok(s):
+        return mt.tokenize(s, return_str=True)
+
+    for line in sys.stdin:
+        parts = list(map(tok, line.split("\t")))
+        print(*parts, sep="\t", flush=True)
+
+
+if __name__ == "__main__":
+    import argparse
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--lang", "-l", default="en")
+    parser.add_argument("--penn", "-p", action="store_true")
+    parser.add_argument("--fields", "-f", help="fields to tokenize")
+    args = parser.parse_args()
+
+    main(args)

fairseq/examples/conv_seq2seq/README.md

@@ -0,0 +1,25 @@
+# Convolutional Sequence to Sequence Learning (Gehring et al., 2017)
+
+## Pre-trained models
+
+Description | Dataset | Model | Test set(s)
+---|---|---|---
+Convolutional <br> ([Gehring et al., 2017](https://arxiv.org/abs/1705.03122)) | [WMT14 English-French](http://statmt.org/wmt14/translation-task.html#Download) | [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/models/wmt14.v2.en-fr.fconv-py.tar.bz2) | newstest2014: <br> [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/data/wmt14.v2.en-fr.newstest2014.tar.bz2) <br> newstest2012/2013: <br> [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/data/wmt14.v2.en-fr.ntst1213.tar.bz2)
+Convolutional <br> ([Gehring et al., 2017](https://arxiv.org/abs/1705.03122)) | [WMT14 English-German](http://statmt.org/wmt14/translation-task.html#Download) | [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/models/wmt14.en-de.fconv-py.tar.bz2) | newstest2014: <br> [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/data/wmt14.en-de.newstest2014.tar.bz2)
+Convolutional <br> ([Gehring et al., 2017](https://arxiv.org/abs/1705.03122)) | [WMT17 English-German](http://statmt.org/wmt17/translation-task.html#Download) | [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/models/wmt17.v2.en-de.fconv-py.tar.bz2) | newstest2014: <br> [download (.tar.bz2)](https://dl.fbaipublicfiles.com/fairseq/data/wmt17.v2.en-de.newstest2014.tar.bz2)
+
+## Example usage
+
+See the [translation README](../translation/README.md) for instructions on reproducing results for WMT'14 En-De and
+WMT'14 En-Fr using the `fconv_wmt_en_de` and `fconv_wmt_en_fr` model architectures.
+
+## Citation
+
+```bibtex
+@inproceedings{gehring2017convs2s,
+  title = {Convolutional Sequence to Sequence Learning},
+  author = {Gehring, Jonas, and Auli, Michael and Grangier, David and Yarats, Denis and Dauphin, Yann N},
+  booktitle = {Proc. of ICML},
+  year = 2017,
+}
+```

fairseq/examples/criss/README.md

@@ -0,0 +1,61 @@
+# Cross-lingual Retrieval for Iterative Self-Supervised Training
+
+https://arxiv.org/pdf/2006.09526.pdf
+
+## Introduction
+
+CRISS is a multilingual sequence-to-sequnce pretraining method where mining and training processes are applied iteratively, improving cross-lingual alignment and translation ability at the same time.
+
+## Requirements:
+
+* faiss: https://github.com/facebookresearch/faiss
+* mosesdecoder: https://github.com/moses-smt/mosesdecoder
+* flores: https://github.com/facebookresearch/flores
+* LASER: https://github.com/facebookresearch/LASER
+
+## Unsupervised Machine Translation
+##### 1. Download and decompress CRISS checkpoints
+```
+cd examples/criss
+wget https://dl.fbaipublicfiles.com/criss/criss_3rd_checkpoints.tar.gz
+tar -xf criss_checkpoints.tar.gz
+```
+##### 2. Download and preprocess Flores test dataset
+Make sure to run all scripts from examples/criss directory
+```
+bash download_and_preprocess_flores_test.sh
+```
+
+##### 3. Run Evaluation on Sinhala-English
+```
+bash unsupervised_mt/eval.sh
+```
+
+## Sentence Retrieval
+##### 1. Download and preprocess Tatoeba dataset
+```
+bash download_and_preprocess_tatoeba.sh
+```
+
+##### 2. Run Sentence Retrieval on Tatoeba Kazakh-English
+```
+bash sentence_retrieval/sentence_retrieval_tatoeba.sh
+```
+
+## Mining
+##### 1. Install faiss
+Follow instructions on https://github.com/facebookresearch/faiss/blob/master/INSTALL.md
+##### 2. Mine pseudo-parallel data between Kazakh and English
+```
+bash mining/mine_example.sh
+```
+
+## Citation
+```bibtex
+@article{tran2020cross,
+  title={Cross-lingual retrieval for iterative self-supervised training},
+  author={Tran, Chau and Tang, Yuqing and Li, Xian and Gu, Jiatao},
+  journal={arXiv preprint arXiv:2006.09526},
+  year={2020}
+}
+```

fairseq/examples/criss/download_and_preprocess_flores_test.sh

@@ -0,0 +1,64 @@
+#!/bin/bash
+# Copyright (c) Facebook, Inc. and its affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+SPM_ENCODE=flores/scripts/spm_encode.py
+DATA=data_tmp
+SPM_MODEL=criss_checkpoints/sentence.bpe.model
+DICT=criss_checkpoints/dict.txt
+
+download_data() {
+  CORPORA=$1
+  URL=$2
+
+  if [ -f $CORPORA ]; then
+    echo "$CORPORA already exists, skipping download"
+  else
+    echo "Downloading $URL"
+    wget $URL -O $CORPORA --no-check-certificate || rm -f $CORPORA
+    if [ -f $CORPORA ]; then
+      echo "$URL successfully downloaded."
+    else
+      echo "$URL not successfully downloaded."
+      rm -f $CORPORA
+    fi
+  fi
+}
+
+if [[ -f flores ]]; then
+  echo "flores already cloned"
+else
+  git clone https://github.com/facebookresearch/flores
+fi
+
+mkdir -p $DATA
+download_data $DATA/wikipedia_en_ne_si_test_sets.tgz "https://github.com/facebookresearch/flores/raw/master/data/wikipedia_en_ne_si_test_sets.tgz"
+pushd $DATA
+pwd
+tar -vxf wikipedia_en_ne_si_test_sets.tgz
+popd
+
+
+for lang in ne_NP si_LK; do
+  datadir=$DATA/${lang}-en_XX-flores
+  rm -rf $datadir
+  mkdir -p $datadir
+  TEST_PREFIX=$DATA/wikipedia_en_ne_si_test_sets/wikipedia.test
+  python $SPM_ENCODE \
+    --model ${SPM_MODEL} \
+    --output_format=piece \
+    --inputs ${TEST_PREFIX}.${lang:0:2}-en.${lang:0:2} ${TEST_PREFIX}.${lang:0:2}-en.en \
+    --outputs $datadir/test.bpe.${lang}-en_XX.${lang} $datadir/test.bpe.${lang}-en_XX.en_XX
+
+  # binarize data
+  fairseq-preprocess \
+    --source-lang ${lang} --target-lang en_XX \
+    --testpref $datadir/test.bpe.${lang}-en_XX \
+    --destdir $datadir \
+    --srcdict ${DICT} \
+    --joined-dictionary \
+    --workers 4
+done

fairseq/examples/criss/download_and_preprocess_tatoeba.sh

@@ -0,0 +1,46 @@
+#!/bin/bash
+# Copyright (c) Facebook, Inc. and its affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+SPM_ENCODE=flores/scripts/spm_encode.py
+DATA=data_tmp
+SPM_MODEL=criss_checkpoints/sentence.bpe.model
+DICT=criss_checkpoints/dict.txt
+
+if [[ -f flores ]]; then
+  echo "flores already cloned"
+else
+  git clone https://github.com/facebookresearch/flores
+fi
+if [[ -f LASER ]]; then
+  echo "LASER already cloned"
+else
+  git clone https://github.com/facebookresearch/LASER
+fi
+mkdir -p data_tmp
+declare -A lang_tatoeba_map=( ["ar_AR"]="ara" ["de_DE"]="deu"  ["es_XX"]="spa" ["et_EE"]="est" ["fi_FI"]="fin" ["fr_XX"]="fra" ["hi_IN"]="hin" ["it_IT"]="ita" ["ja_XX"]="jpn" ["ko_KR"]="kor" ["kk_KZ"]="kaz" ["nl_XX"]="nld" ["ru_RU"]="rus" ["tr_TR"]="tur" ["vi_VN"]="vie" ["zh_CN"]="cmn")
+for lang in ar_AR de_DE es_XX et_EE fi_FI fr_XX hi_IN it_IT ja_XX kk_KZ ko_KR nl_XX ru_RU tr_TR vi_VN zh_CN; do
+  lang_tatoeba=${lang_tatoeba_map[$lang]}
+  echo $lang_tatoeba
+  datadir=$DATA/${lang}-en_XX-tatoeba
+  rm -rf $datadir
+  mkdir -p $datadir
+  TEST_PREFIX=LASER/data/tatoeba/v1/tatoeba
+  python $SPM_ENCODE \
+    --model ${SPM_MODEL} \
+    --output_format=piece \
+    --inputs ${TEST_PREFIX}.${lang_tatoeba}-eng.${lang_tatoeba} ${TEST_PREFIX}.${lang_tatoeba}-eng.eng \
+    --outputs $datadir/test.bpe.${lang}-en_XX.${lang} $datadir/test.bpe.${lang}-en_XX.en_XX
+
+  # binarize data
+  fairseq-preprocess \
+    --source-lang ${lang} --target-lang en_XX \
+    --testpref $datadir/test.bpe.${lang}-en_XX \
+    --destdir $datadir \
+    --srcdict ${DICT} \
+    --joined-dictionary \
+    --workers 4
+done

fairseq/examples/criss/mining/mine.py

@@ -0,0 +1,240 @@
+#!/usr/bin/env python3 -u
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+import argparse
+import glob
+from subprocess import check_call
+
+try:
+    import faiss
+
+    has_faiss = True
+except ImportError:
+    has_faiss = False
+import numpy as np
+
+
+GB = 1024 * 1024 * 1024
+
+
+def call(cmd):
+    print(cmd)
+    check_call(cmd, shell=True)
+
+
+def get_batches(directory, lang, prefix="all_avg_pool"):
+    print(f"Finding in {directory}/{prefix}.{lang}*")
+    files = glob.glob(f"{directory}/{prefix}.{lang}*")
+    emb_files = []
+    txt_files = []
+    for emb_fi in files:
+        emb_files.append(emb_fi)
+        txt_fi = emb_fi.replace(prefix, "sentences")
+        txt_files.append(txt_fi)
+    return emb_files, txt_files
+
+
+def load_batch(emb_file, dim):
+    embeddings = np.fromfile(emb_file, dtype=np.float32)
+    num_rows = int(embeddings.shape[0] / dim)
+    embeddings = embeddings.reshape((num_rows, dim))
+    faiss.normalize_L2(embeddings)
+    return embeddings
+
+
+def knnGPU_sharded(x_batches_f, y_batches_f, dim, k, direction="x2y"):
+    if not has_faiss:
+        raise ImportError("Please install Faiss")
+    sims = []
+    inds = []
+    xfrom = 0
+    xto = 0
+    for x_batch_f in x_batches_f:
+        yfrom = 0
+        yto = 0
+        x_batch = load_batch(x_batch_f, dim)
+        xto = xfrom + x_batch.shape[0]
+        bsims, binds = [], []
+        for y_batch_f in y_batches_f:
+            y_batch = load_batch(y_batch_f, dim)
+            neighbor_size = min(k, y_batch.shape[0])
+            yto = yfrom + y_batch.shape[0]
+            print("{}-{}  ->  {}-{}".format(xfrom, xto, yfrom, yto))
+            idx = faiss.IndexFlatIP(dim)
+            idx = faiss.index_cpu_to_all_gpus(idx)
+            idx.add(y_batch)
+            bsim, bind = idx.search(x_batch, neighbor_size)
+
+            bsims.append(bsim)
+            binds.append(bind + yfrom)
+            yfrom += y_batch.shape[0]
+            del idx
+            del y_batch
+        bsims = np.concatenate(bsims, axis=1)
+        binds = np.concatenate(binds, axis=1)
+        aux = np.argsort(-bsims, axis=1)
+        sim_batch = np.zeros((x_batch.shape[0], k), dtype=np.float32)
+        ind_batch = np.zeros((x_batch.shape[0], k), dtype=np.int64)
+        for i in range(x_batch.shape[0]):
+            for j in range(k):
+                sim_batch[i, j] = bsims[i, aux[i, j]]
+                ind_batch[i, j] = binds[i, aux[i, j]]
+        sims.append(sim_batch)
+        inds.append(ind_batch)
+        xfrom += x_batch.shape[0]
+        del x_batch
+    sim = np.concatenate(sims, axis=0)
+    ind = np.concatenate(inds, axis=0)
+    return sim, ind
+
+
+def score(sim, fwd_mean, bwd_mean, margin):
+    return margin(sim, (fwd_mean + bwd_mean) / 2)
+
+
+def score_candidates(
+    sim_mat, candidate_inds, fwd_mean, bwd_mean, margin, verbose=False
+):
+    print(" - scoring {:d} candidates".format(sim_mat.shape[0]))
+    scores = np.zeros(candidate_inds.shape)
+    for i in range(scores.shape[0]):
+        for j in range(scores.shape[1]):
+            k = int(candidate_inds[i, j])
+            scores[i, j] = score(sim_mat[i, j], fwd_mean[i], bwd_mean[k], margin)
+    return scores
+
+
+def load_text(files):
+    all_sentences = []
+    for fi in files:
+        with open(fi) as sentence_fi:
+            for line in sentence_fi:
+                all_sentences.append(line.strip())
+    print(f"Read {len(all_sentences)} sentences")
+    return all_sentences
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Mine bitext")
+    parser.add_argument("--src-lang", help="Source language")
+    parser.add_argument("--tgt-lang", help="Target language")
+    parser.add_argument(
+        "--dict-path", help="Path to dictionary file", default="dict.txt"
+    )
+    parser.add_argument(
+        "--spm-path", help="Path to SPM model file", default="sentence.bpe.model"
+    )
+    parser.add_argument("--dim", type=int, default=1024, help="Embedding dimension")
+    parser.add_argument("--mem", type=int, default=5, help="Memory in GB")
+    parser.add_argument("--src-dir", help="Source directory")
+    parser.add_argument("--tgt-dir", help="Target directory")
+    parser.add_argument("--output", help="Output path")
+    parser.add_argument(
+        "--neighborhood", type=int, default=4, help="Embedding dimension"
+    )
+    parser.add_argument(
+        "--threshold", type=float, default=1.06, help="Threshold on mined bitext"
+    )
+    parser.add_argument(
+        "--valid-size",
+        type=int,
+        default=2000,
+        help="Number of sentences used for validation set",
+    )
+    parser.add_argument(
+        "--min-count",
+        type=int,
+        default=50000,
+        help="Min num sentences used for each language",
+    )
+    args = parser.parse_args()
+
+    x_batches_f, x_sents_f = get_batches(args.src_dir, args.src_lang)
+    y_batches_f, y_sents_f = get_batches(args.tgt_dir, args.tgt_lang)
+    margin = lambda a, b: a / b
+    y2x_sim, y2x_ind = knnGPU_sharded(
+        y_batches_f, x_batches_f, args.dim, args.neighborhood, direction="y2x"
+    )
+    x2y_sim, x2y_ind = knnGPU_sharded(
+        x_batches_f, y_batches_f, args.dim, args.neighborhood, direction="x2y"
+    )
+
+    x2y_mean = x2y_sim.mean(axis=1)
+    y2x_mean = y2x_sim.mean(axis=1)
+    fwd_scores = score_candidates(x2y_sim, x2y_ind, x2y_mean, y2x_mean, margin)
+    bwd_scores = score_candidates(y2x_sim, y2x_ind, y2x_mean, x2y_mean, margin)
+    fwd_best = x2y_ind[np.arange(x2y_sim.shape[0]), fwd_scores.argmax(axis=1)]
+    bwd_best = y2x_ind[np.arange(y2x_sim.shape[0]), bwd_scores.argmax(axis=1)]
+    indices = np.stack(
+        (
+            np.concatenate((np.arange(x2y_ind.shape[0]), bwd_best)),
+            np.concatenate((fwd_best, np.arange(y2x_ind.shape[0]))),
+        ),
+        axis=1,
+    )
+    scores = np.concatenate((fwd_scores.max(axis=1), bwd_scores.max(axis=1)))
+
+    x_sentences = load_text(x_sents_f)
+    y_sentences = load_text(y_sents_f)
+
+    threshold = args.threshold
+    min_count = args.min_count
+    seen_src, seen_trg = set(), set()
+    directory = args.output
+    call(f"mkdir -p {directory}")
+    src_out = open(
+        f"{directory}/all.{args.src_lang}",
+        mode="w",
+        encoding="utf-8",
+        errors="surrogateescape",
+    )
+    tgt_out = open(
+        f"{directory}/all.{args.tgt_lang}",
+        mode="w",
+        encoding="utf-8",
+        errors="surrogateescape",
+    )
+    scores_out = open(
+        f"{directory}/all.scores", mode="w", encoding="utf-8", errors="surrogateescape"
+    )
+    count = 0
+    for i in np.argsort(-scores):
+        src_ind, trg_ind = indices[i]
+        if src_ind not in seen_src and trg_ind not in seen_trg:
+            seen_src.add(src_ind)
+            seen_trg.add(trg_ind)
+            if scores[i] > threshold or count < min_count:
+                if x_sentences[src_ind]:
+                    print(scores[i], file=scores_out)
+                    print(x_sentences[src_ind], file=src_out)
+                    print(y_sentences[trg_ind], file=tgt_out)
+                    count += 1
+                else:
+                    print(f"Ignoring sentence: {x_sentences[src_ind]}")
+    src_out.close()
+    tgt_out.close()
+    scores_out.close()
+
+    print(f"Found {count} pairs for threshold={threshold}")
+    with open(f"{directory}/all.{args.src_lang}") as all_s, open(
+        f"{directory}/all.{args.tgt_lang}"
+    ) as all_t, open(f"{directory}/valid.{args.src_lang}", "w") as valid_s, open(
+        f"{directory}/valid.{args.tgt_lang}", "w"
+    ) as valid_t, open(
+        f"{directory}/train.{args.src_lang}", "w"
+    ) as train_s, open(
+        f"{directory}/train.{args.tgt_lang}", "w"
+    ) as train_t:
+        count = 0
+        for s_line, t_line in zip(all_s, all_t):
+            s_line = s_line.split("\t")[1]
+            t_line = t_line.split("\t")[1]
+            if count >= args.valid_size:
+                train_s.write(s_line)
+                train_t.write(t_line)
+            else:
+                valid_s.write(s_line)
+                valid_t.write(t_line)
+                count += 1

fairseq/examples/criss/mining/mine_example.sh

@@ -0,0 +1,103 @@
+#!/bin/bash
+# Copyright (c) Facebook, Inc. and its affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+#
+source_lang=kk_KZ
+target_lang=en_XX
+MODEL=criss_checkpoints/criss.3rd.pt
+SPM=criss_checkpoints/sentence.bpe.model
+SPLIT=test
+LANG_DICT=criss_checkpoints/lang_dict.txt
+SPM_ENCODE=flores/scripts/spm_encode.py
+SAVE_ENCODER=save_encoder.py
+ENCODER_SAVE_ROOT=sentence_embeddings/$MODEL
+DICT=criss_checkpoints/dict.txt
+THRESHOLD=1.02
+MIN_COUNT=500
+
+DATA_DIR=data_tmp
+SAVE_DIR=mining/${source_lang}_${target_lang}_mined
+ENCODER_SAVE_DIR=${ENCODER_SAVE_ROOT}/${source_lang}-${target_lang}
+INPUT_DIR=$DATA_DIR/${source_lang}-${target_lang}-tatoeba
+
+mkdir -p $ENCODER_SAVE_DIR/${target_lang}
+mkdir -p $ENCODER_SAVE_DIR/${source_lang}
+mkdir -p $SAVE_DIR
+
+## Save encoder outputs
+
+# Save encoder outputs for source sentences
+python $SAVE_ENCODER \
+  ${INPUT_DIR} \
+  --path ${MODEL} \
+  --task translation_multi_simple_epoch \
+  --lang-pairs ${source_lang}-${target_lang} \
+  --lang-dict ${LANG_DICT} \
+  --gen-subset ${SPLIT} \
+  --bpe 'sentencepiece' \
+  -s ${source_lang} -t ${target_lang} \
+  --sentencepiece-model ${SPM} \
+  --remove-bpe 'sentencepiece' \
+  --beam 1 \
+  --lang-tok-style mbart \
+  --encoder-save-dir ${ENCODER_SAVE_DIR}/${source_lang}
+
+## Save encoder outputs for target sentences
+python $SAVE_ENCODER \
+  ${INPUT_DIR} \
+  --path ${MODEL} \
+  --lang-pairs ${source_lang}-${target_lang} \
+  --lang-dict ${LANG_DICT} \
+  --task translation_multi_simple_epoch \
+  --gen-subset ${SPLIT} \
+  --bpe 'sentencepiece' \
+  -t ${source_lang} -s ${target_lang} \
+  --sentencepiece-model ${SPM} \
+  --remove-bpe 'sentencepiece' \
+  --beam 1 \
+  --lang-tok-style mbart \
+  --encoder-save-dir ${ENCODER_SAVE_DIR}/${target_lang}
+
+## Mining
+python mining/mine.py \
+  --src-lang ${source_lang} \
+  --tgt-lang ${target_lang} \
+  --dim 1024 \
+  --mem 10 \
+  --neighborhood 4 \
+  --src-dir ${ENCODER_SAVE_DIR}/${source_lang} \
+  --tgt-dir ${ENCODER_SAVE_DIR}/${target_lang} \
+  --output $SAVE_DIR \
+  --threshold ${THRESHOLD} \
+  --min-count ${MIN_COUNT} \
+  --valid-size 100 \
+  --dict-path ${DICT} \
+  --spm-path ${SPM} \
+
+
+## Process and binarize mined data
+python $SPM_ENCODE \
+  --model ${SPM} \
+  --output_format=piece \
+  --inputs mining/${source_lang}_${target_lang}_mined/train.${source_lang} mining/${source_lang}_${target_lang}_mined/train.${target_lang} \
+  --outputs mining/${source_lang}_${target_lang}_mined/train.bpe.${source_lang} mining/${source_lang}_${target_lang}_mined/train.bpe.${target_lang}
+
+python $SPM_ENCODE \
+  --model ${SPM} \
+  --output_format=piece \
+  --inputs mining/${source_lang}_${target_lang}_mined/valid.${source_lang} mining/${source_lang}_${target_lang}_mined/valid.${target_lang} \
+  --outputs mining/${source_lang}_${target_lang}_mined/valid.bpe.${source_lang} mining/${source_lang}_${target_lang}_mined/valid.bpe.${target_lang}
+
+
+fairseq-preprocess \
+  --source-lang ${source_lang} \
+  --target-lang ${target_lang} \
+  --trainpref mining/${source_lang}_${target_lang}_mined/train.bpe \
+  --validpref mining/${source_lang}_${target_lang}_mined/valid.bpe \
+  --destdir mining/${source_lang}_${target_lang}_mined \
+  --srcdict ${DICT} \
+  --joined-dictionary \
+  --workers 8

fairseq/examples/criss/save_encoder.py

@@ -0,0 +1,214 @@
+#!/usr/bin/env python3 -u
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+"""
+Translate pre-processed data with a trained model.
+"""
+
+import numpy as np
+import torch
+from fairseq import checkpoint_utils, options, progress_bar, tasks, utils
+from fairseq.sequence_generator import EnsembleModel
+from fairseq.utils import safe_hasattr
+
+
+def get_avg_pool(
+    models, sample, prefix_tokens, src_dict, remove_bpe, has_langtok=False
+):
+    model = EnsembleModel(models)
+
+    # model.forward normally channels prev_output_tokens into the decoder
+    # separately, but SequenceGenerator directly calls model.encoder
+    encoder_input = {
+        k: v for k, v in sample["net_input"].items() if k != "prev_output_tokens"
+    }
+
+    # compute the encoder output for each beam
+    encoder_outs = model.forward_encoder(encoder_input)
+    np_encoder_outs = encoder_outs[0].encoder_out.cpu().numpy().astype(np.float32)
+    encoder_mask = 1 - encoder_outs[0].encoder_padding_mask.cpu().numpy().astype(
+        np.float32
+    )
+    encoder_mask = np.expand_dims(encoder_mask.T, axis=2)
+    if has_langtok:
+        encoder_mask = encoder_mask[1:, :, :]
+        np_encoder_outs = np_encoder_outs[1, :, :]
+    masked_encoder_outs = encoder_mask * np_encoder_outs
+    avg_pool = (masked_encoder_outs / encoder_mask.sum(axis=0)).sum(axis=0)
+    return avg_pool
+
+
+def main(args):
+    assert args.path is not None, "--path required for generation!"
+    assert (
+        not args.sampling or args.nbest == args.beam
+    ), "--sampling requires --nbest to be equal to --beam"
+    assert (
+        args.replace_unk is None or args.raw_text
+    ), "--replace-unk requires a raw text dataset (--raw-text)"
+
+    args.beam = 1
+    utils.import_user_module(args)
+
+    if args.max_tokens is None:
+        args.max_tokens = 12000
+    print(args)
+    use_cuda = torch.cuda.is_available() and not args.cpu
+
+    # Load dataset splits
+    task = tasks.setup_task(args)
+    task.load_dataset(args.gen_subset)
+
+    # Set dictionaries
+    try:
+        src_dict = getattr(task, "source_dictionary", None)
+    except NotImplementedError:
+        src_dict = None
+    tgt_dict = task.target_dictionary
+
+    # Load ensemble
+    print("| loading model(s) from {}".format(args.path))
+    models, _model_args = checkpoint_utils.load_model_ensemble(
+        args.path.split(":"),
+        arg_overrides=eval(args.model_overrides),
+        task=task,
+    )
+
+    # Optimize ensemble for generation
+    for model in models:
+        model.make_generation_fast_(
+            beamable_mm_beam_size=None if args.no_beamable_mm else args.beam,
+            need_attn=args.print_alignment,
+        )
+        if args.fp16:
+            model.half()
+        if use_cuda:
+            model.cuda()
+
+    # Load alignment dictionary for unknown word replacement
+    # (None if no unknown word replacement, empty if no path to align dictionary)
+    align_dict = utils.load_align_dict(args.replace_unk)
+
+    # Load dataset (possibly sharded)
+    itr = task.get_batch_iterator(
+        dataset=task.dataset(args.gen_subset),
+        max_tokens=args.max_tokens,
+        max_positions=utils.resolve_max_positions(
+            task.max_positions(),
+        ),
+        ignore_invalid_inputs=args.skip_invalid_size_inputs_valid_test,
+        required_batch_size_multiple=args.required_batch_size_multiple,
+        num_shards=args.num_shards,
+        shard_id=args.shard_id,
+        num_workers=args.num_workers,
+    ).next_epoch_itr(shuffle=False)
+
+    num_sentences = 0
+    source_sentences = []
+    shard_id = 0
+    all_avg_pool = None
+    encoder_has_langtok = (
+        safe_hasattr(task.args, "encoder_langtok")
+        and task.args.encoder_langtok is not None
+        and safe_hasattr(task.args, "lang_tok_replacing_bos_eos")
+        and not task.args.lang_tok_replacing_bos_eos
+    )
+    with progress_bar.build_progress_bar(args, itr) as t:
+        for sample in t:
+            if sample is None:
+                print("Skipping None")
+                continue
+            sample = utils.move_to_cuda(sample) if use_cuda else sample
+            if "net_input" not in sample:
+                continue
+
+            prefix_tokens = None
+            if args.prefix_size > 0:
+                prefix_tokens = sample["target"][:, : args.prefix_size]
+
+            with torch.no_grad():
+                avg_pool = get_avg_pool(
+                    models,
+                    sample,
+                    prefix_tokens,
+                    src_dict,
+                    args.post_process,
+                    has_langtok=encoder_has_langtok,
+                )
+                if all_avg_pool is not None:
+                    all_avg_pool = np.concatenate((all_avg_pool, avg_pool))
+                else:
+                    all_avg_pool = avg_pool
+
+            if not isinstance(sample["id"], list):
+                sample_ids = sample["id"].tolist()
+            else:
+                sample_ids = sample["id"]
+            for i, sample_id in enumerate(sample_ids):
+                # Remove padding
+                src_tokens = utils.strip_pad(
+                    sample["net_input"]["src_tokens"][i, :], tgt_dict.pad()
+                )
+
+                # Either retrieve the original sentences or regenerate them from tokens.
+                if align_dict is not None:
+                    src_str = task.dataset(args.gen_subset).src.get_original_text(
+                        sample_id
+                    )
+                else:
+                    if src_dict is not None:
+                        src_str = src_dict.string(src_tokens, args.post_process)
+                    else:
+                        src_str = ""
+
+                if not args.quiet:
+                    if src_dict is not None:
+                        print("S-{}\t{}".format(sample_id, src_str))
+
+                source_sentences.append(f"{sample_id}\t{src_str}")
+
+            num_sentences += sample["nsentences"]
+            if all_avg_pool.shape[0] >= 1000000:
+                with open(
+                    f"{args.encoder_save_dir}/all_avg_pool.{args.source_lang}.{shard_id}",
+                    "w",
+                ) as avg_pool_file:
+                    all_avg_pool.tofile(avg_pool_file)
+                with open(
+                    f"{args.encoder_save_dir}/sentences.{args.source_lang}.{shard_id}",
+                    "w",
+                ) as sentence_file:
+                    sentence_file.writelines(f"{line}\n" for line in source_sentences)
+                all_avg_pool = None
+                source_sentences = []
+                shard_id += 1
+
+    if all_avg_pool is not None:
+        with open(
+            f"{args.encoder_save_dir}/all_avg_pool.{args.source_lang}.{shard_id}", "w"
+        ) as avg_pool_file:
+            all_avg_pool.tofile(avg_pool_file)
+        with open(
+            f"{args.encoder_save_dir}/sentences.{args.source_lang}.{shard_id}", "w"
+        ) as sentence_file:
+            sentence_file.writelines(f"{line}\n" for line in source_sentences)
+    return None
+
+
+def cli_main():
+    parser = options.get_generation_parser()
+    parser.add_argument(
+        "--encoder-save-dir",
+        default="",
+        type=str,
+        metavar="N",
+        help="directory to save encoder outputs",
+    )
+    args = options.parse_args_and_arch(parser)
+    main(args)
+
+
+if __name__ == "__main__":
+    cli_main()

fairseq/examples/criss/sentence_retrieval/encoder_analysis.py

@@ -0,0 +1,92 @@
+#!/usr/bin/env python3 -u
+# Copyright (c) Facebook, Inc. and its affiliates.
+#
+# This source code is licensed under the MIT license found in the
+# LICENSE file in the root directory of this source tree.
+import argparse
+import glob
+
+import numpy as np
+
+
+DIM = 1024
+
+
+def compute_dist(source_embs, target_embs, k=5, return_sim_mat=False):
+    target_ids = [tid for tid in target_embs]
+    source_mat = np.stack(source_embs.values(), axis=0)
+    normalized_source_mat = source_mat / np.linalg.norm(
+        source_mat, axis=1, keepdims=True
+    )
+    target_mat = np.stack(target_embs.values(), axis=0)
+    normalized_target_mat = target_mat / np.linalg.norm(
+        target_mat, axis=1, keepdims=True
+    )
+    sim_mat = normalized_source_mat.dot(normalized_target_mat.T)
+    if return_sim_mat:
+        return sim_mat
+    neighbors_map = {}
+    for i, sentence_id in enumerate(source_embs):
+        idx = np.argsort(sim_mat[i, :])[::-1][:k]
+        neighbors_map[sentence_id] = [target_ids[tid] for tid in idx]
+    return neighbors_map
+
+
+def load_embeddings(directory, LANGS):
+    sentence_embeddings = {}
+    sentence_texts = {}
+    for lang in LANGS:
+        sentence_embeddings[lang] = {}
+        sentence_texts[lang] = {}
+        lang_dir = f"{directory}/{lang}"
+        embedding_files = glob.glob(f"{lang_dir}/all_avg_pool.{lang}.*")
+        for embed_file in embedding_files:
+            shard_id = embed_file.split(".")[-1]
+            embeddings = np.fromfile(embed_file, dtype=np.float32)
+            num_rows = embeddings.shape[0] // DIM
+            embeddings = embeddings.reshape((num_rows, DIM))
+
+            with open(f"{lang_dir}/sentences.{lang}.{shard_id}") as sentence_file:
+                for idx, line in enumerate(sentence_file):
+                    sentence_id, sentence = line.strip().split("\t")
+                    sentence_texts[lang][sentence_id] = sentence
+                    sentence_embeddings[lang][sentence_id] = embeddings[idx, :]
+
+    return sentence_embeddings, sentence_texts
+
+
+def compute_accuracy(directory, LANGS):
+    sentence_embeddings, sentence_texts = load_embeddings(directory, LANGS)
+
+    top_1_accuracy = {}
+
+    top1_str = " ".join(LANGS) + "\n"
+    for source_lang in LANGS:
+        top_1_accuracy[source_lang] = {}
+        top1_str += f"{source_lang} "
+        for target_lang in LANGS:
+            top1 = 0
+            top5 = 0
+            neighbors_map = compute_dist(
+                sentence_embeddings[source_lang], sentence_embeddings[target_lang]
+            )
+            for sentence_id, neighbors in neighbors_map.items():
+                if sentence_id == neighbors[0]:
+                    top1 += 1
+                if sentence_id in neighbors[:5]:
+                    top5 += 1
+            n = len(sentence_embeddings[target_lang])
+            top1_str += f"{top1/n} "
+        top1_str += "\n"
+
+    print(top1_str)
+    print(top1_str, file=open(f"{directory}/accuracy", "w"))
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Analyze encoder outputs")
+    parser.add_argument("directory", help="Source language corpus")
+    parser.add_argument("--langs", help="List of langs")
+    args = parser.parse_args()
+    langs = args.langs.split(",")
+    compute_accuracy(args.directory, langs)