diff options
author | Taku Kudo <taku910@users.noreply.github.com> | 2018-05-01 13:14:31 +0300 |
---|---|---|
committer | GitHub <noreply@github.com> | 2018-05-01 13:14:31 +0300 |
commit | c0d9a5d2630530c9395c784b7f2c1c21e06bb7f4 (patch) | |
tree | a61a38c71e52dbb59395801a780bc11485982849 /README.md | |
parent | c18b5949cc6c3ee7ee5b2a0ba0f6c5f65d9d3715 (diff) |
Update README.md
Diffstat (limited to 'README.md')
-rw-r--r-- | README.md | 2 |
1 files changed, 1 insertions, 1 deletions
@@ -89,7 +89,7 @@ Subword regularization [[Kudo.](https://arxiv.org/abs/1804.10959)] is a simple r that virtually augments training data with on-the-fly subword sampling, which helps to improve the accuracy as well as robustness of NMT models. To enable subword regularization, you would like to use the SentencePiece library -([C++](doc/api.md)/[Python](python/README.md)) to sample one segmentation for each parameter updates, which is different from the standard off-line data preparations. Here's the example of [Python library](python/README.md). You can find that 'New York' is segmented differently on each ``SampleEncode`` call. The details of sampling parameters are found in [sentencepiece_processor.h](src/sentencepiece_processor.h). +([C++](doc/api.md#sampling-subword-regularization)/[Python](python/README.md)) to sample one segmentation for each parameter updates, which is different from the standard off-line data preparations. Here's the example of [Python library](python/README.md). You can find that 'New York' is segmented differently on each ``SampleEncode`` call. The details of sampling parameters are found in [sentencepiece_processor.h](src/sentencepiece_processor.h). ``` >>> import sentencepiece as spm |