|
|
Switch to news-commentary-v12
|
 Sergey Edunov
|
|
6 years ago |
|
|
Prepare scripts for WMT14
|
Sergey Edunov
|
|
6 years ago |
|
|
Directly decay weight instead of L2 penalty (#157)
|
 Yann N. Dauphin
|
|
6 years ago |
|
|
Merge pull request #77 from facebookresearch/oss-merge-internal
|
 Myle Ott
|
|
6 years ago |
|
|
Fix conv padding for even kernel widths
|
Myle Ott
|
|
6 years ago |
|
|
Rebuild optimizer when loading checkpoints
|
Myle Ott
|
|
6 years ago |
|
|
Fix weight norm dimension in decoder (fixes #73)
|
Myle Ott
|
|
6 years ago |
|
|
Save number of GPUs in args (and checkpoints)
|
Myle Ott
|
|
6 years ago |
|
|
Prefer command-line configuration over checkpoint for optimizer state
|
Myle Ott
|
|
6 years ago |
|
|
Allow --lr to specify a fixed learning rate schedule
|
Myle Ott
|
|
6 years ago |
|
|
Improve memory handling (recover from OOM and periodically empty caching allocator)
|
Myle Ott
|
|
6 years ago |
|
|
Improve error when resuming training with a different model architecture
|
Myle Ott
|
|
6 years ago |
|
|
Fixed 2 typos (#75)
|
 toothlessdragon
|
|
6 years ago |
|
|
fix bug in lstm model (#68)
|
 Zrachel
|
|
6 years ago |
|
|
Update requirements.txt and fix flake8 (#62)
|
Myle Ott
|
|
6 years ago |
|
|
Remove Python3.6 format string from preprocess.py (fixes #60) (#61)
|
Myle Ott
|
|
6 years ago |
|
|
Remove more Python 3.6 format strings (fixes #57) (#58)
|
Myle Ott
|
|
6 years ago |
|
|
Merge pull request #56 from facebookresearch/python3
|
Myle Ott
|
|
6 years ago |
|
|
Remove Python 3.6 format strings (fixes #55)
|
Myle Ott
|
|
6 years ago |
|
|
Make LSTM backwards compatible and fix incremental generation
|
Myle Ott
|
|
6 years ago |
|
|
Flush non-TTY logging output after each log interval
|
Myle Ott
|
|
6 years ago |
|
|
Fix Flake8
|
Myle Ott
|
|
6 years ago |
|
|
Fallback to `--log-format=simple` for non-TTY terminals
|
Myle Ott
|
|
6 years ago |
|
|
Merge pull request #54: Version 0.1.0 -> 0.2.0
|
Myle Ott
|
|
6 years ago |
|
|
Version 0.1.0 -> 0.2.0
|
Myle Ott
|
|
6 years ago |
|
|
Fix all-reduce for new versions of PyTorch
|
Myle Ott
|
|
6 years ago |
|
|
Fixes for `--log-format`
|
Myle Ott
|
|
6 years ago |
|
|
Fix max_positions_valid in train.py
|
Myle Ott
|
|
6 years ago |
|
|
Add `--log-format` option and JSON logger
|
Myle Ott
|
|
6 years ago |
|
|
Don't call forward directly (prefer module(x) to module.forward(x))
|
Myle Ott
|
|
6 years ago |
Connect to our Discord channel
Models
Annotations