From f9e15ad025f117b38cf03d3b2c75628c4202c0ed Mon Sep 17 00:00:00 2001
From: Yigit Sever
Date: Sat, 21 Sep 2019 14:55:33 +0300
Subject: Clean up demo.sh

---
 demo.sh | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

(limited to 'demo.sh')

diff --git a/demo.sh b/demo.sh
index 990a106..d8b0b37 100755
--- a/demo.sh
+++ b/demo.sh
@@ -37,6 +37,7 @@ EMBS="${ROOT}/embeddings"
 MAP_TO="${ROOT}/bilingual_embeddings"
 
 # create wordnets directory and download a single wordnet
+echo "Downloading one wordnet"
 mkdir -p "${WNET}"
 wget -nc -q http://compling.hss.ntu.edu.sg/omw/wns/bul.zip -P "${WNET}"
 unzip -o -q "${WNET}/bul.zip" -d "${WNET}"
@@ -46,11 +47,13 @@ mkdir -p "${TAB_DIR}"
 "${SCRIPTS}/tab_creator.pl" "${WNET}/bul/wn-data-bul.tab" "${TAB_DIR}"
 
 # create ready directory and create two .def files
+echo "Creating two .def files"
 mkdir -p "${READY}"
 python "${SCRIPTS}/prep_lookup.py" -s "en" -t "bg"
 mv "${ROOT}"/*.def "${READY}"
 
 # create dictionaries directory and download a single dictionary
+echo "Creating seed lexicons"
 mkdir -p "${DICT}"
 wget -nc -q https://object.pouta.csc.fi/OPUS-OpenSubtitles/v2018/dic/bg-en.dic.gz -P "${DICT}" # Bulgarian - English
 gunzip -q "${DICT}/bg-en.dic.gz"
@@ -67,6 +70,7 @@ mv "${DICT}"/*.test "${TEST_DIR}"
 rm -f "${DICT}"/*.dic
 
 # download two monolingual embeddings
+echo "Downloading monolingual embeddings"
 wget -nc -q https://dl.fbaipublicfiles.com/fasttext/vectors-crawl/cc.bg.300.vec.gz -P "${EMBS}" # Bulgarian
 wget -nc -q https://dl.fbaipublicfiles.com/fasttext/vectors-english/crawl-300d-2M.vec.zip -P "${EMBS}" # English
 gunzip "${EMBS}/cc.bg.300.vec.gz"
@@ -80,21 +84,15 @@ for lang_code in bg en; do
     sed -i '1 s/^.*$/500000 300/' "${EMBS}/${lang_code}.vec"
 done
 
-# map two embeddings
-python "${ROOT}/vecmap/map_embeddings.py" --supervised \
-    "${TRAIN_DIC_DIR}/en_bg.train" \
-    "${EMBS}/en.vec" \
-    "${EMBS}/bg.vec" \
-    "${MAP_TO}/en_to_bg.vec" \
-    "${MAP_TO}/bg_to_en.vec" > /dev/null 2>&1
-
+echo "Mapping bilingual embeddings"
 mkdir -p "${MAP_TO}" # create bilingual embeddings directory
 source_lang="en"
 target_lang="bg"
-
 python "${ROOT}/vecmap/map_embeddings.py" --supervised \
-    "${TRAIN_DIC_DIR}/${source_lang}_${target_lang}.train" \
+    "${TRAIN_DIR}/${source_lang}_${target_lang}.train" \
     "${EMBS}/${source_lang}.vec" \
     "${EMBS}/${target_lang}.vec" \
     "${MAP_TO}/${source_lang}_to_${target_lang}.vec" \
     "${MAP_TO}/${target_lang}_to_${source_lang}.vec" > /dev/null 2>&1
+
+python "${ROOT}/WMD_matching.py" "${source_lang}" "${target_lang}" "${MAP_TO}/${source_lang}_to_${target_lang}.vec" "${MAP_TO}/${target_lang}_to_${source_lang}.vec" "${READY}/${source_lang}_to_${target_lang}.def" "${READY}/${target_lang}_to_${source_lang}.def" all
-- 
cgit v1.2.3-70-g09d2