sacremoses Documentation

Documentation

Open alvations opened this issue 6 years ago • 5 comments

Now that we have more than tokenization, we need some proper documentation.

Dec 18 '18 10:12 alvations

Need to hurry up with the docs...

ETA: 20 May 19

Sorry for the delay for whoever is following this repo ;P

Jan 25 '19 06:01 alvations

Bumping issue -_- |||

Please do this asap...

Apr 13 '20 04:04 alvations

I'd like to try sacremoses because I am in need of a truecaser. The documentation does not link to a pretrained model. Where can I find it?

May 04 '20 20:05 mfoglio

@mfoglio there are no pre-trained models in this library. They are purely rule-based ;P

May 05 '20 00:05 alvations

Ok, so just to be sure I expressed myself correctly. I see in the home that you can train MosesTruecaser().

For instance: mtr = MosesTruecaser() mtr.truecase("hello my friend")

Returns:

AssertionError: 
Use Truecaser.train() to train a model.
Or use Truecaser('modefile') to load a model.

Where can I find the rule-based models?

May 05 '20 15:05 mfoglio