From 243db8dff4171056dabed8bf81b3422bde675487 Mon Sep 17 00:00:00 2001 From: Yigit Sever Date: Mon, 23 Sep 2019 01:47:07 +0300 Subject: Add demo instructions and # of pairs --- README.md | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) diff --git a/README.md b/README.md index 4c5073e..8de096e 100644 --- a/README.md +++ b/README.md @@ -66,6 +66,28 @@ Then; This will create two directories; `dictionaries` and `wordnets`. Linewise aligned definition files are in `wordnets/ready`. +
Language pairs and number of available aligned glosses +

+Source Language | Target Language | # of Pairs +--- | --- | ---: +en | bg | 4959 +en | el | 18136 +en | it | 12688 +en | ro | 58754 +en | sl | 3144 +en | sq | 4681 +bg | el | 2817 +bg | it | 2115 +bg | ro | 4701 +el | it | 4801 +el | ro | 2144 +el | sq | 4681 +it | ro | 10353 +ro | sl | 2085 +ro | sq | 4646 +

+
+ ## Acquiring The Embeddings We use [VecMap](https://github.com/artetxem/vecmap) on [fastText](https://fasttext.cc/) embeddings. You can skip this step if you are providing your own polylingual embeddings. @@ -88,3 +110,10 @@ git submodule init && git submodule update Bear in mind that this will require around 50 GB free space. +## Quick Demo + +`demo.sh` is included, downloads data for 2 languages. + +```bash +./demo.sh +``` -- cgit v1.2.3-70-g09d2