|
allow specifying max_tokens for generation
|
Alexei Baevski
|
|
6 years ago |
|
remove completed sentences from batch
|
Alexei Baevski
|
|
6 years ago |
|
Merge internal changes (#136)
|
Myle Ott
|
|
6 years ago |
|
Allow more flexible pre-processing and generation (#227)
|
Sergey Edunov
|
|
6 years ago |
|
More fixes for recent PyTorch (incl. topk issue) (#113)
|
Myle Ott
|
|
6 years ago |
|
Add support to prefixes (#221)
|
Dario Pavllo
|
|
6 years ago |
|
More unit test fixes
|
Myle Ott
|
|
6 years ago |
|
fairseq-py goes distributed (#106)
|
Myle Ott
|
|
6 years ago |
|
Support deprecation of volatile Variables in latest PyTorch
|
Myle Ott
|
|
6 years ago |
|
Add support for sharded generation
|
Myle Ott
|
|
6 years ago |
|
Fallback to `--log-format=simple` for non-TTY terminals
|
Myle Ott
|
|
6 years ago |
|
Add `--log-format` option and JSON logger
|
Myle Ott
|
|
6 years ago |
|
Replace unk with original string
|
Louis Martin
|
|
6 years ago |
|
Add --max-sentence option for batching based on # sentences
|
Myle Ott
|
|
6 years ago |
|
Update README with interactive.py and fix it
|
Louis Martin
|
|
6 years ago |
|
Fix interactive.py
|
Myle Ott
|
|
6 years ago |
|
Improvements to data loader
|
Myle Ott
|
|
6 years ago |
|
Refactor generation
|
Louis Martin
|
|
6 years ago |
|
Added -unkpen flag to generate.py following logic of Lua/Torch version
|
Michael Auli
|
|
6 years ago |
|
Refactor model definitions
|
Myle Ott
|
|
6 years ago |
|
Fix handling of continuation tokens that precede <unk> in generate.py
|
Myle Ott
|
|
6 years ago |
|
Simplify deps of build_model to only depend on dict (instead of dataset)
|
Myle Ott
|
|
6 years ago |
|
Fix flake8 warnings
|
Louis Martin
|
|
6 years ago |
|
Support configurable BPE symbol
|
Myle Ott
|
|
6 years ago |
|
Move helper functions from generate.py to fairseq/dictionary.py
|
Myle Ott
|
|
6 years ago |
|
Ignore invalid sentences in test and valid
|
Sergey Edunov
|
|
6 years ago |
|
Don't generate during training, add --quiet to generate.py
|
Sergey Edunov
|
|
6 years ago |
|
Update progress_bar to be more robust to changes in tqdm (#21)
|
Myle Ott
|
|
6 years ago |
|
Fix generation when vocabulary is small relative to beam size (fixes #7)
|
Myle Ott
|
|
6 years ago |
|
More fixes
|
Sergey Edunov
|
|
6 years ago |