GitLab

Summary:
The previous BSD+PATENTS license was controversial. We have been
approved to relicense fairseq under the MIT license.
Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/786

Differential Revision: D16560654

Pulled By: myleott

fbshipit-source-id: f78b1beb4f2895dd7b9bfc79f5f952a2bfb94034

Summary:
Implement Nucleus (top-P) sampling: sample among the smallest set of elements whose cumulative probability mass exceeds p.

To test it:
python generate.py   ~myleott/data/data-bin/wmt17_zh_en_full/   --path ~myleott/zh_en/model.pt   --remove-bpe   --nbest 5   --beam 5 --sampling --sampling-topp 0.3
Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/710

Test Plan:
python generate.py   ~myleott/data/data-bin/wmt17_zh_en_full/   --path ~myleott/zh_en/model.pt   --remove-bpe   --nbest 5   --beam 5 --sampling --sampling-topp 0.3

python tests/test_sequence_generator.py

python tests/test_binaries.py

Reviewed By: myleott

Differential Revision: D16286688

Pulled By: xingz9

fbshipit-source-id: 1776d21e17c4532a3d24ac75bb7e75da9acad58f

Summary:
Pull Request resolved: https://github.com/pytorch/translate/pull/351

This makes it easier for tasks to plugin to generate.py/interactive.py
Pull Request resolved: https://github.com/pytorch/fairseq/pull/520

Differential Revision: D14183881

Pulled By: myleott

fbshipit-source-id: ede5e53ddc1215ed3b12b8f1eba048c946913c33

Co-authored-by: liezl200 <lie@fb.com>

A Task defines the data format, stores shared state (e.g., dictionaries) and provides helpers for building the model/criterion and calculating the loss.

Changes:
- Add TranslationTask and LanguageModelingTask. New tasks can be registered with @register_task decorator.
- Add EpochBatchIterator to encapsulate batching and saving/restoring dataloader position
- Remove LEFT_PAD_* constants and make them configurable per task

Changes:
- 7d19e36: Add `--sampling` flag to generate.py to sample instead of doing beam search
- c777340: Add `scripts/average_checkpoints.py` to average multiple checkpoints into a combined model
- 3ea882c: Add `--max-update` option to train.py to stop training after a given number of updates
- small bugfixes for distributed training, LSTM, inverse square root LR scheduler

This PR includes breaking API changes to modularize fairseq-py and adds support for distributed training across multiple nodes.

Changes:
- c7033ef: add support for distributed training! See updated README for usage.
- e016299: modularize fairseq-py, adding support for register_model, register_criterion, register_optimizer, etc.
- 154e440: update LSTM implementation to use PackedSequence objects in the encoder, better following best practices and improving perf
- 90c2973 and 1da6265: improve unit test coverage