diff --git a/.gitattributes b/.gitattributes
index c59b0739094913e90ec71efd49e74c9e6f4cb840..c046b1fe0e5878bf6ffeba04451c8129bd3b9e35 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -143,3 +143,237 @@ data/data.su.json filter=lfs diff=lfs merge=lfs -text
 data/data.yue.json filter=lfs diff=lfs merge=lfs -text
 data/data.lmo.json filter=lfs diff=lfs merge=lfs -text
 data/data.oc.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.gd.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.pt.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sl.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.lmo.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.mr.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.yi.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ilo.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ko.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.tl.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.uz.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ar.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.eml.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.hy.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ia.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ml.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sv.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.zh.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.als.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ckb.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ga.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.mt.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.uk.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.vi.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.am.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.cy.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.fr.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.lv.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.bg.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.dv.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.hi.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.rm.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.fa.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.is.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ja.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.la.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.mn.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.tg.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ug.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.af.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.fi.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.kn.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.mk.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sa.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.su.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ta.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ps.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.be.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.kk.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.km.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.or.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.es.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.et.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.si.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.cs.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ky.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.th.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ka.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sq.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sw.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.hr.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ht.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.vep.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sd.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.lb.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ne.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.da.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.io.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ku.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.lt.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.fy.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.te.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.hu.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.yue.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.no.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.arz.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.jv.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.lo.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.mg.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.jbo.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.nn.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sh.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ceb.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.gl.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.it.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sk.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.el.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.gu.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ms.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.pnb.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ur.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.as.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.az.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.de.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.eo.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ro.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.bs.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.eu.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.id.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.pl.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.nds.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.tr.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ca.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.en.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.he.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.my.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ce.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.sr.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.ru.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.so.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.tt.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.war.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.wuu.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.br.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.nl.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.oc.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.pa.json filter=lfs diff=lfs merge=lfs -text
+data/train/data.bn.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sh.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.so.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.be.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.da.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.km.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.kn.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.no.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.or.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.su.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.as.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ceb.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.cy.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.id.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.kk.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.tg.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.arz.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ca.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ka.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.cs.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.hu.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.mg.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ru.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.jbo.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.la.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.bn.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.de.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ar.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.lv.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sv.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.yi.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.et.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.eu.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.io.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ku.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ml.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ps.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sd.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.th.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.uz.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.war.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.lmo.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.si.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.dv.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.es.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.lo.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.hy.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.lt.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sw.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.als.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ce.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.lb.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.my.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sk.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ja.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.oc.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sr.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.pnb.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.te.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ckb.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.fa.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.pa.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ky.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.wuu.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ga.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.mt.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sl.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.tl.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ht.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ko.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.mn.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.nds.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ta.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.br.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.eo.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ia.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.mk.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.yue.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.bs.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.gu.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.jv.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.vi.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.zh.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.el.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.vep.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.am.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.az.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.tt.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.en.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.gd.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.gl.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.nl.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sq.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.af.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.rm.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.sa.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ug.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.tr.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.fi.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.fy.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ilo.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.is.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ms.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.pl.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.it.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.fr.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.hi.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.mr.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.nn.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.pt.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.eml.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ne.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ro.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.uk.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.ur.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.bg.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.he.json filter=lfs diff=lfs merge=lfs -text
+data/valid/data.hr.json filter=lfs diff=lfs merge=lfs -text
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000000000000000000000000000000000000..503304d496e2927ddc370ed90dfe9693df0ef5d1
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+data/*.json
\ No newline at end of file
diff --git a/data/train/data.af.json b/data/train/data.af.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e81ea1a098d74efb90815cdc3e16e6abbd60387
--- /dev/null
+++ b/data/train/data.af.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0316a7d42165f0251cb9be866644291aac70352752b038fed208645b6e98586
+size 198915
diff --git a/data/train/data.als.json b/data/train/data.als.json
new file mode 100644
index 0000000000000000000000000000000000000000..634a803632fe6edeff52d9ed3d09c455d3dfcbea
--- /dev/null
+++ b/data/train/data.als.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6aacce34104e965e2aa11a45f3a2a2fa87dc7e46679c8ee449f1ccfdccabbdd
+size 653
diff --git a/data/train/data.am.json b/data/train/data.am.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6b3d435c4b29e85cee4a51ec68cf6ecbf440b6a
--- /dev/null
+++ b/data/train/data.am.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28b1d46ee248a78327e860eb63cbf14513672321de0e40bd2b7a9ad81c7b92bf
+size 65608
diff --git a/data/train/data.ar.json b/data/train/data.ar.json
new file mode 100644
index 0000000000000000000000000000000000000000..a709f566a10ddabbc21235275bd35c864e68047f
--- /dev/null
+++ b/data/train/data.ar.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5ede355e54e23a747c3161b8ed34e432fe90d3f419d5486a35d6f22d17f1d48
+size 59984865
diff --git a/data/train/data.arz.json b/data/train/data.arz.json
new file mode 100644
index 0000000000000000000000000000000000000000..c234b274de3ecc70fc71e19608d5e53f08c18b21
--- /dev/null
+++ b/data/train/data.arz.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dc81c51b92dfd3b465ba6a0a04c0b9b791e03a180ca87d277e58230f9ca6c1c
+size 3712
diff --git a/data/train/data.as.json b/data/train/data.as.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e96f5c3066e2ded8b8859f2d5d545a6665f05ce
--- /dev/null
+++ b/data/train/data.as.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8620a00ddee6ed68b84ca2cecc31c8c9f77001ad5c82d0f3736b22a388f7586c
+size 10748
diff --git a/data/train/data.az.json b/data/train/data.az.json
new file mode 100644
index 0000000000000000000000000000000000000000..e12d56cd4aa7d25a9e8c7b5b849ee8b8becb393d
--- /dev/null
+++ b/data/train/data.az.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd663bb241037fce68915862f1d07207dd1dc41f73aa4e3a7ed64af917f4de99
+size 915031
diff --git a/data/train/data.be.json b/data/train/data.be.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7acfc7df47b37e1610860a67d721bfba7883a4d
--- /dev/null
+++ b/data/train/data.be.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2378a87a4bc294a03be51fe028d87a75dcb303e1a02f6f4c37e29755235d81aa
+size 116146
diff --git a/data/train/data.bg.json b/data/train/data.bg.json
new file mode 100644
index 0000000000000000000000000000000000000000..740fd4efdf8a1b2b367503efa312a7d8f6cb56eb
--- /dev/null
+++ b/data/train/data.bg.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f0186ea6ebdf825c55e2914a614d5ce39af01c15d3f11e8d6e826fbd81b5455
+size 34077793
diff --git a/data/train/data.bn.json b/data/train/data.bn.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd6386d2d8622c78e4cb5affa384872c0e9e457d
--- /dev/null
+++ b/data/train/data.bn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5970a058e7b9387600be0e90ca9a1be9f2757a00103c09dc1dacea7b424fd0d
+size 5053092
diff --git a/data/train/data.br.json b/data/train/data.br.json
new file mode 100644
index 0000000000000000000000000000000000000000..46312fd36c4748f1a8a9142e53282a30f59bef90
--- /dev/null
+++ b/data/train/data.br.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a4ec258078f5ffa226627d96236d6fea24935e698d856beb6234f2ef7d6f2dd
+size 216
diff --git a/data/train/data.bs.json b/data/train/data.bs.json
new file mode 100644
index 0000000000000000000000000000000000000000..750c4eda9e97200d85f674306d1867bd57329d40
--- /dev/null
+++ b/data/train/data.bs.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92fc76facc4c465291b7a9884a13a5f72462ee0191485d60ee0e84225c73df30
+size 49666
diff --git a/data/train/data.ca.json b/data/train/data.ca.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfb9bf146bc9b6a0948a5cac5e550fda26e2a618
--- /dev/null
+++ b/data/train/data.ca.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7ecd075efbc5bb08e99a356ccf4942eefba6bc0cdcda477bf9876a6b92c158f
+size 6293281
diff --git a/data/train/data.ce.json b/data/train/data.ce.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e101c5eb9b908bc9a58362daf1f6d75af32f488
--- /dev/null
+++ b/data/train/data.ce.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b7c9a4e60467163f7ee4cbc20941f93e25d63e5fa21208d589703d446875f92
+size 1103
diff --git a/data/train/data.ceb.json b/data/train/data.ceb.json
new file mode 100644
index 0000000000000000000000000000000000000000..17c46015f9c8841ffe1e9d680670f5edbce8e4ab
--- /dev/null
+++ b/data/train/data.ceb.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e177f17c57c395684858dbc928c220dd98ee595d2d63a98b2d083b4f8a594cc
+size 994563
diff --git a/data/train/data.ckb.json b/data/train/data.ckb.json
new file mode 100644
index 0000000000000000000000000000000000000000..73ee643fe5d3d96484a82041784bb3837c0a6696
--- /dev/null
+++ b/data/train/data.ckb.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6372eb141665e99481f00841fcbcb1d53767eb07976efe08ed7aa0fbc593ee99
+size 49570
diff --git a/data/train/data.cs.json b/data/train/data.cs.json
new file mode 100644
index 0000000000000000000000000000000000000000..3376aacaf967d8843ec880a20cc67288d7b194a4
--- /dev/null
+++ b/data/train/data.cs.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbe99fb994e1e7cb22b18d28afa47b1046dc4bc538d9c571d210d90aea80bdda
+size 32453990
diff --git a/data/train/data.cy.json b/data/train/data.cy.json
new file mode 100644
index 0000000000000000000000000000000000000000..273b31f6c2dc2f72157e3765b90890b48639b2a9
--- /dev/null
+++ b/data/train/data.cy.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf5349adf5ee956aac8a1a8d0b79711f9184c394b5f0f44c70876e0f7d5dba14
+size 374807
diff --git a/data/train/data.da.json b/data/train/data.da.json
new file mode 100644
index 0000000000000000000000000000000000000000..50b7bfbbb1983bf3d3a6ceed14e4684d0c99f18b
--- /dev/null
+++ b/data/train/data.da.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bbaa638b12f17edf9f89001fa90f5652b4dc64cabf3b22dcd1113ae45f8f837
+size 53402481
diff --git a/data/train/data.de.json b/data/train/data.de.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5376c04e167d3dc8feecd8a2ff37d3cec7d0491
--- /dev/null
+++ b/data/train/data.de.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdf2703481f395d53f33c19e09293366b002954520d4150a3e93381dc2c9bd35
+size 647104170
diff --git a/data/train/data.dv.json b/data/train/data.dv.json
new file mode 100644
index 0000000000000000000000000000000000000000..a951dd1f6690d4ae4e56b793c680b89112b5dd77
--- /dev/null
+++ b/data/train/data.dv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bb1192b6c6d4f00aa0fe48daf3ee065805e256f3de39cbd2a9b2feeb6decbeb
+size 12432
diff --git a/data/train/data.el.json b/data/train/data.el.json
new file mode 100644
index 0000000000000000000000000000000000000000..70b1a374d0357801488036120ce6c2c5f1e2726e
--- /dev/null
+++ b/data/train/data.el.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78f26bad31241dd2a0d69673354266c54ee3a14c16727574edf8f4643ff4fa70
+size 40708463
diff --git a/data/train/data.eml.json b/data/train/data.eml.json
new file mode 100644
index 0000000000000000000000000000000000000000..565a08dcc79a4560c6798aa3640f5e6e1ad770f5
--- /dev/null
+++ b/data/train/data.eml.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e6fcde4f5b5982256eedaf1bf955c14c93b954db194ab774d749facd1cbb99a
+size 345
diff --git a/data/train/data.en.json b/data/train/data.en.json
new file mode 100644
index 0000000000000000000000000000000000000000..795552c897dfc1e43db09622b2a3ef0b8aec18df
--- /dev/null
+++ b/data/train/data.en.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdce76d7893f927534bb649351375dfcb689e97bf8df340266a8455d7b1657a4
+size 3580537410
diff --git a/data/train/data.eo.json b/data/train/data.eo.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0ea007c7fdd1e309ed73ff4c7079babef651bd2
--- /dev/null
+++ b/data/train/data.eo.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb0c0a4c1b86c02a69c97338bff8a9a120d86058c3fd75af3103fc53a5555271
+size 96806
diff --git a/data/train/data.es.json b/data/train/data.es.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f2d3e76a9f659d1e25da6fef5615a42687a144c
--- /dev/null
+++ b/data/train/data.es.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e19a787e6653211f0e6237de5aa06abaf75dfa8a60d7ff12dc87f8e9db29a44d
+size 355480187
diff --git a/data/train/data.et.json b/data/train/data.et.json
new file mode 100644
index 0000000000000000000000000000000000000000..c4f15e42c4121fe78739d9112ff1b248f69e28af
--- /dev/null
+++ b/data/train/data.et.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65872e4d1e738e403c77881096eda2e52e899cbe16b1b030166410d40ba57c9b
+size 9185759
diff --git a/data/train/data.eu.json b/data/train/data.eu.json
new file mode 100644
index 0000000000000000000000000000000000000000..78497fbd88f5c1b0fa787ef4a3218b39a556e071
--- /dev/null
+++ b/data/train/data.eu.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78253ade75713849c428d24b7bd9fc1ec2cdbcebcc75dc32e61363a1cf841c9b
+size 292251
diff --git a/data/train/data.fa.json b/data/train/data.fa.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab1a1a0a49f6fe21cc260f49165b258a132bd8cd
--- /dev/null
+++ b/data/train/data.fa.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de8fcd252f7b3e9b52e56b2e24bfc8665c341c8cdfd1c4ae11b5a0a793d17f95
+size 193594051
diff --git a/data/train/data.fi.json b/data/train/data.fi.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac8efefc7d0867afff7981501c4e204aac8928a0
--- /dev/null
+++ b/data/train/data.fi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96ea2fdbe15b76784705bf7530cf3a9fe65c4f4cedd7cc415039e9dad2e6a08b
+size 36140450
diff --git a/data/train/data.fr.json b/data/train/data.fr.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c2068f10fe60e2dcf7b2e82b35b9d525cc12154
--- /dev/null
+++ b/data/train/data.fr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0c0a2b8a9848974ace5d0033c2a2c67cc5f9c1846e56b3ded1c808386fe5eee
+size 384307467
diff --git a/data/train/data.fy.json b/data/train/data.fy.json
new file mode 100644
index 0000000000000000000000000000000000000000..a02606865dc9894654873315e09f7eacf7f69edd
--- /dev/null
+++ b/data/train/data.fy.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:757a70650695d1fc06eb5187ee9fdab7a35ed44cb15268c195321a8b5275ef8e
+size 22050
diff --git a/data/train/data.ga.json b/data/train/data.ga.json
new file mode 100644
index 0000000000000000000000000000000000000000..303d847d5d6e23e12d7dbcece5d6e0aa01576498
--- /dev/null
+++ b/data/train/data.ga.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:637b31d6ca6355583a0191eab8331dacf73c4b57349805b6697b12d8d578915a
+size 301196
diff --git a/data/train/data.gd.json b/data/train/data.gd.json
new file mode 100644
index 0000000000000000000000000000000000000000..0938dfe28ed5c41a41f304e3ff19cf61431ad0ba
--- /dev/null
+++ b/data/train/data.gd.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ebcd272d72b298265bd4f8b0d39f2881a3e56339f1ce3ce964f1f7927b5f5a9
+size 49151
diff --git a/data/train/data.gl.json b/data/train/data.gl.json
new file mode 100644
index 0000000000000000000000000000000000000000..cd8ce272a7bbd21ee73a24c3d59493f63656cb52
--- /dev/null
+++ b/data/train/data.gl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0fae90b0b31f9b117fcfce1a88c2c63f3519b924b688581f5e35b78ad93b7de
+size 114434
diff --git a/data/train/data.gu.json b/data/train/data.gu.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a604cdf38c2fbda2b0409d0a1c8d20a322e7581
--- /dev/null
+++ b/data/train/data.gu.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc6151b9e71f0c3dc23ba83a8364ef1f48fd244ddb031b46dd985c9f78e8b6cb
+size 623244
diff --git a/data/train/data.he.json b/data/train/data.he.json
new file mode 100644
index 0000000000000000000000000000000000000000..53563c41a34f1fe61d44542dfb64f8cf99708473
--- /dev/null
+++ b/data/train/data.he.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6bf90449e406791d092f4ad3a11f607722c193f02c3081993bcd7dea72fa559
+size 50632822
diff --git a/data/train/data.hi.json b/data/train/data.hi.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7ce9776d14bc4c706ef961a50b08086a06290cf
--- /dev/null
+++ b/data/train/data.hi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:080c6f8a28910a78306a64bb3ff973b8cfce4e789cadb69f0da5d7e96fc84cfd
+size 45880519
diff --git a/data/train/data.hr.json b/data/train/data.hr.json
new file mode 100644
index 0000000000000000000000000000000000000000..75a877efd9da3ba2d87f0d8cb07e2c97f4d778f6
--- /dev/null
+++ b/data/train/data.hr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9756d1c12c57894638ee1025a11406d378b6c6c47a93dbd7be85fb4cc8065fd
+size 10479255
diff --git a/data/train/data.ht.json b/data/train/data.ht.json
new file mode 100644
index 0000000000000000000000000000000000000000..69caa0881315961bef94436ada7e0759f248aca7
--- /dev/null
+++ b/data/train/data.ht.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c02175b2c4c29962322559769ea23a646b1c25044240057c29de6d8135e3255b
+size 62569
diff --git a/data/train/data.hu.json b/data/train/data.hu.json
new file mode 100644
index 0000000000000000000000000000000000000000..735772404aeae36ee781b357742838ccb4e02e48
--- /dev/null
+++ b/data/train/data.hu.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17509fec8c944f3bc5a20ad3bd385ae011c17d5710b683d4e3be750a0eba2e60
+size 30655722
diff --git a/data/train/data.hy.json b/data/train/data.hy.json
new file mode 100644
index 0000000000000000000000000000000000000000..91547da9c5919d029fd890ea9f48a12b711cfc1d
--- /dev/null
+++ b/data/train/data.hy.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e31193a32c4e5b4d608833551d55afad0e9d15a102b63f68da31a5f6eb9cf571
+size 407894
diff --git a/data/train/data.ia.json b/data/train/data.ia.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ca310ebc39e1697fcec82cbd793e2befc2592c6
--- /dev/null
+++ b/data/train/data.ia.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1968f38d8b0532c5542296630e4dba10c2e541e3d8b641db732e247af09c6ff
+size 1615
diff --git a/data/train/data.id.json b/data/train/data.id.json
new file mode 100644
index 0000000000000000000000000000000000000000..354dff3b25455ef151baea874d9d363999452461
--- /dev/null
+++ b/data/train/data.id.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99f189dbd20b85ae225276e50647f82bf503b99c48fd6098b11f47dda8586798
+size 52735229
diff --git a/data/train/data.ilo.json b/data/train/data.ilo.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c39d12fba861b393a9517b731ac703b54646a07
--- /dev/null
+++ b/data/train/data.ilo.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90fef6ac13e7df84ffb6a707bff3bb7137aecf692dc74460468f77a0cda72d33
+size 1192
diff --git a/data/train/data.io.json b/data/train/data.io.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd0e2bcd458d6b2a331dd5abe0c0c99f31b47aa9
--- /dev/null
+++ b/data/train/data.io.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:587a221d14eeb2de218f4a35cc092044d2d9f1cac890bdc13b9462457959b848
+size 199
diff --git a/data/train/data.is.json b/data/train/data.is.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf57769fcfdedfd02bd6a4ddfc22b8a4cda4fb92
--- /dev/null
+++ b/data/train/data.is.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3870a71cc60cc616dbcea9d0f80158997d5e625eea63c62378449d55a43c58e
+size 1429004
diff --git a/data/train/data.it.json b/data/train/data.it.json
new file mode 100644
index 0000000000000000000000000000000000000000..383544df7cec13aa84f2a12df8d130703ddd1f3a
--- /dev/null
+++ b/data/train/data.it.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31d5c1d0c1c1a6378ba82430319a038d881cab642383c08088575a5e83ccbf7f
+size 157857500
diff --git a/data/train/data.ja.json b/data/train/data.ja.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cd024e03f836cb4a05417e0e7c5d5f50715e4bc
--- /dev/null
+++ b/data/train/data.ja.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdfea2661848be333ebba4df413aaf133e7d6f95d08e93b7640237026cec0802
+size 188564307
diff --git a/data/train/data.jbo.json b/data/train/data.jbo.json
new file mode 100644
index 0000000000000000000000000000000000000000..6e30845fef28185e69b90d297942ee91b680b2f8
--- /dev/null
+++ b/data/train/data.jbo.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d2eb2ee37de5473c3b2a1c38ec0d5164d1ae2c9591952ca20c3e88783ab2daa
+size 1856
diff --git a/data/train/data.jv.json b/data/train/data.jv.json
new file mode 100644
index 0000000000000000000000000000000000000000..de0ffeb4ebdac60af5e09feea9909ba02bd4c4ef
--- /dev/null
+++ b/data/train/data.jv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdc732df6cdd34f89c17b09e83ee9cf12f5bd3091844dac05ae70854c754be6c
+size 13181
diff --git a/data/train/data.ka.json b/data/train/data.ka.json
new file mode 100644
index 0000000000000000000000000000000000000000..57cfdbcc852f16a07b8f62d890fbda0b35ea7882
--- /dev/null
+++ b/data/train/data.ka.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbf609df0c89bfda39b0ed6e26cd096da18b806d92a664563e233e30085aa475
+size 1599506
diff --git a/data/train/data.kk.json b/data/train/data.kk.json
new file mode 100644
index 0000000000000000000000000000000000000000..39abbbdadc3b3980f16314b065fdf7d195eaa9dd
--- /dev/null
+++ b/data/train/data.kk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acac7d5aab040a6fbc80480c5bc6687d82a17997ff16e545a5a38d27f07268a0
+size 902992
diff --git a/data/train/data.km.json b/data/train/data.km.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb08495efc77d874f6372317b4a00fa115e6c6dd
--- /dev/null
+++ b/data/train/data.km.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78ed9a495ffedace5e244552e2d6f8a103cf886babe6824a18ce74da967cc44f
+size 309977
diff --git a/data/train/data.kn.json b/data/train/data.kn.json
new file mode 100644
index 0000000000000000000000000000000000000000..84d159f01dec4dbb0e5c0b44f60df17dfd3bf9cc
--- /dev/null
+++ b/data/train/data.kn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f49052d78fc81d823b55e6217df6e151f0f27aa45b08783c83e86e43ef504748
+size 396841
diff --git a/data/train/data.ko.json b/data/train/data.ko.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e773ba1cd3dc81f5f42cdc1621c024f6d126b25
--- /dev/null
+++ b/data/train/data.ko.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4aabd52cc76758a05828c057c796b31c7b7b63762a4034aa74ce0cfb3b29367
+size 19530797
diff --git a/data/train/data.ku.json b/data/train/data.ku.json
new file mode 100644
index 0000000000000000000000000000000000000000..c9242b18c590018bb9372b53130fb53ee3eb5cf9
--- /dev/null
+++ b/data/train/data.ku.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:002f741d5c3d3b5d8640b7dcb3040bf89a8a534c7fe3867405282d31a769c0db
+size 43657
diff --git a/data/train/data.ky.json b/data/train/data.ky.json
new file mode 100644
index 0000000000000000000000000000000000000000..130513528bc3d380e904a8155c9c03cb0c2db845
--- /dev/null
+++ b/data/train/data.ky.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:304df36e25a3e6b0f3d254c92f7fda86f3a1163a76b25f1ebfef2495baba7319
+size 218221
diff --git a/data/train/data.la.json b/data/train/data.la.json
new file mode 100644
index 0000000000000000000000000000000000000000..6c9e407623e1b6faf0fd3143103c2d01fad13402
--- /dev/null
+++ b/data/train/data.la.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c548ac0a55c0e9289c308520eee00050a13f31b9dd5974bb43d827bb689bd479
+size 1671751
diff --git a/data/train/data.lb.json b/data/train/data.lb.json
new file mode 100644
index 0000000000000000000000000000000000000000..74cd1fa647d6da76a359131dbaae3eb795c99987
--- /dev/null
+++ b/data/train/data.lb.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207238e443b5018ee433327bd84cbb30665f7ca09f6bdf386771e05fb3aa95fd
+size 199277
diff --git a/data/train/data.lmo.json b/data/train/data.lmo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4153db073ef50b0aa1a57481ae90378db3b89de
--- /dev/null
+++ b/data/train/data.lmo.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9c01f8cede0d6f63d9e7fb809c28571f3e24e31e4184fd2795fe5c7d3e9888c
+size 275
diff --git a/data/train/data.lo.json b/data/train/data.lo.json
new file mode 100644
index 0000000000000000000000000000000000000000..b0110f4381f255eb3f9379e4c75396d0457b3be9
--- /dev/null
+++ b/data/train/data.lo.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6c6ab35d2d79599bf016bdd36d749ce9661bf4c4b4b51a7660a6c5c8709b404
+size 333530
diff --git a/data/train/data.lt.json b/data/train/data.lt.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc4146c9e78ca02a4dee8e8b2d8d98884c107a70
--- /dev/null
+++ b/data/train/data.lt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed75794218d34e4a3eb854e1bf7e813edf7993312437504909c67f0c26ae1d8c
+size 8825486
diff --git a/data/train/data.lv.json b/data/train/data.lv.json
new file mode 100644
index 0000000000000000000000000000000000000000..514ee8ddd1a3f4082b212b5d6b9eb15043f92a3d
--- /dev/null
+++ b/data/train/data.lv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:283c105be7cb93c3276cdde4df055b0a10a6c958c7bbaf730107153a537c57f7
+size 5545538
diff --git a/data/train/data.mg.json b/data/train/data.mg.json
new file mode 100644
index 0000000000000000000000000000000000000000..85c98211bbe5027b79725b5fa44ea3cbd3acc0a0
--- /dev/null
+++ b/data/train/data.mg.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5487f446e12492f119a2698874e318da6643db186f8bc02298c4aaf8dde2f12f
+size 66062
diff --git a/data/train/data.mk.json b/data/train/data.mk.json
new file mode 100644
index 0000000000000000000000000000000000000000..98daf5c1f78d1c12988dfe47cfadaeb814ecd721
--- /dev/null
+++ b/data/train/data.mk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:798dd2b15ebae9e289e366ee0a76bda40e754e7c171db28c939c78abfe2c503b
+size 522812
diff --git a/data/train/data.ml.json b/data/train/data.ml.json
new file mode 100644
index 0000000000000000000000000000000000000000..9aea0842ceff33c1e13826eacec75d7a94e9cb43
--- /dev/null
+++ b/data/train/data.ml.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34e07a3e94b51511e0e0f41bf96dd17e1bd8c43d550edede7929f627353475d8
+size 475415
diff --git a/data/train/data.mn.json b/data/train/data.mn.json
new file mode 100644
index 0000000000000000000000000000000000000000..f0c69de5c5c6e407411d574215d1d4d13eb80343
--- /dev/null
+++ b/data/train/data.mn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4652ebe9e4452b4ed0ce6f10d368fb5441eec0d1842b24675a05105d04bbf2b
+size 203697
diff --git a/data/train/data.mr.json b/data/train/data.mr.json
new file mode 100644
index 0000000000000000000000000000000000000000..899fe9a5de10f44dd33aff232637750429fdb5fc
--- /dev/null
+++ b/data/train/data.mr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08cb6d235653d975761b99da45206b7a5ce941f88d99222a43b4318b34710bd9
+size 1333178
diff --git a/data/train/data.ms.json b/data/train/data.ms.json
new file mode 100644
index 0000000000000000000000000000000000000000..34949b8db1ed1dade24ed82fd138229a8ab82e20
--- /dev/null
+++ b/data/train/data.ms.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a5de93262a02c1ab436e2c9c07a67efa6bca3369b7281105499636f37bebb8c
+size 1752912
diff --git a/data/train/data.mt.json b/data/train/data.mt.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fbb1284240f31edaab3bb03097764f66f311d72
--- /dev/null
+++ b/data/train/data.mt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2127ac1698533ba34513d53fb82ca8c5792bad174df525007d97a8bc1b67056c
+size 348005
diff --git a/data/train/data.my.json b/data/train/data.my.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f0330d1a7f381af438a212cba75658754ccc502
--- /dev/null
+++ b/data/train/data.my.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6436382cf57194ce470034d4d7f7c4135dcf8f43bec0247ef71fb0148f4a0bf
+size 797604
diff --git a/data/train/data.nds.json b/data/train/data.nds.json
new file mode 100644
index 0000000000000000000000000000000000000000..abe5784a174c25c17c7937316f0acbc147d38188
--- /dev/null
+++ b/data/train/data.nds.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28d7694f75755e1a1404bbd65db4af9df1cd56aeb71af960e025d69c044a8370
+size 8588
diff --git a/data/train/data.ne.json b/data/train/data.ne.json
new file mode 100644
index 0000000000000000000000000000000000000000..be2813b148e890fec38ef9a4efa12db3c6baff05
--- /dev/null
+++ b/data/train/data.ne.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41bfe070f44b694cdbf6d941d125d8e7f563af15bfe0c21184194b5147cbebf5
+size 253039
diff --git a/data/train/data.nl.json b/data/train/data.nl.json
new file mode 100644
index 0000000000000000000000000000000000000000..b704f34a258ce0234c961d06c5775078a5e6828f
--- /dev/null
+++ b/data/train/data.nl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa59cab3776009ba9014e40ed0f936454567c4c22942dadae5922ca6285becde
+size 260516690
diff --git a/data/train/data.nn.json b/data/train/data.nn.json
new file mode 100644
index 0000000000000000000000000000000000000000..428382f9b53b05d2bf774d11f6d2df9d9e26e4c2
--- /dev/null
+++ b/data/train/data.nn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:398b27d04d08fd7bf22aeb773c04150565db12f87e099eec831b1987bd77e8b8
+size 39695
diff --git a/data/train/data.no.json b/data/train/data.no.json
new file mode 100644
index 0000000000000000000000000000000000000000..78456bfda7e76cd355201a5a853ed29f962c5424
--- /dev/null
+++ b/data/train/data.no.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a86f2b0d04715915b0883db804a2326dea3a98652ac1ae6f97d4e1e2a476808
+size 29915018
diff --git a/data/train/data.oc.json b/data/train/data.oc.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a69d2b3bf27ecbc95d79194ec9c8045df472212
--- /dev/null
+++ b/data/train/data.oc.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d11df70774d0d23315052d49995cc168b371faf0890f9fc4e8d115ce61f2dc9
+size 1409
diff --git a/data/train/data.or.json b/data/train/data.or.json
new file mode 100644
index 0000000000000000000000000000000000000000..022108ca3a4a1b0e83d9db8cfd22cd6cdc5b88a5
--- /dev/null
+++ b/data/train/data.or.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c20aa59aeaaed4765b72c8dca66c3e0c9af2116be5d244b04e109267c0d308b1
+size 6164
diff --git a/data/train/data.pa.json b/data/train/data.pa.json
new file mode 100644
index 0000000000000000000000000000000000000000..9cefd32d1a458adcbbde0fd34ee2ded4330c8acd
--- /dev/null
+++ b/data/train/data.pa.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4538d70a5c805537b28d9a2f3f2ba22729d044802101ea7929bafc1a0160b8d
+size 205339
diff --git a/data/train/data.pl.json b/data/train/data.pl.json
new file mode 100644
index 0000000000000000000000000000000000000000..835c5540466bcd9ba91480946ac2cfea2dd2c77a
--- /dev/null
+++ b/data/train/data.pl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b78f6359a4485ef28ebd8f187708c7171f45d70943582aa15b9be43e54694ab
+size 109200747
diff --git a/data/train/data.pnb.json b/data/train/data.pnb.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b73d108bd2021b5206bae9a160b4ec1100ea0fe
--- /dev/null
+++ b/data/train/data.pnb.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d572c66058d70127b7c7b6c7f76d20d9249d7033b50bb2f03af6276fb252439b
+size 1054
diff --git a/data/train/data.ps.json b/data/train/data.ps.json
new file mode 100644
index 0000000000000000000000000000000000000000..491cfe86d02150b1fa02e63dff9a779e58cc4063
--- /dev/null
+++ b/data/train/data.ps.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4be6dc8b736fc9eaf5e9fa676c50ea8ec9a6d03b3fc5b90f34a7690f8cc60aa
+size 135025
diff --git a/data/train/data.pt.json b/data/train/data.pt.json
new file mode 100644
index 0000000000000000000000000000000000000000..07c3f9c7d11d4c34d0c580eab076e00af6cbc4d5
--- /dev/null
+++ b/data/train/data.pt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06aaf895697332f5fe6dc6c4a00c6eeac50b3d0bc8cabe97a5d8adc071a49d81
+size 124588355
diff --git a/data/train/data.rm.json b/data/train/data.rm.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e491d157ad958265779f979a3c1d33eb1f86660
--- /dev/null
+++ b/data/train/data.rm.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbdc65763cce0bdb7b31b87d8b0ec5481fcfc04220d6223cf3d23c52e8c23eeb
+size 276
diff --git a/data/train/data.ro.json b/data/train/data.ro.json
new file mode 100644
index 0000000000000000000000000000000000000000..da9ae857dca25c3206d987bad6b5763ed55ec292
--- /dev/null
+++ b/data/train/data.ro.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbed6acbc346664ba882a0633f7d039a55643cdb136401dd70a92ee405308980
+size 31534574
diff --git a/data/train/data.ru.json b/data/train/data.ru.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ccc6014e81004264efc4c5682e8583c91114d61
--- /dev/null
+++ b/data/train/data.ru.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7edaf3ff0b3f94dc78ed01b737c87ea885ad3e18b25f539775f961ea6edc8399
+size 600990553
diff --git a/data/train/data.sa.json b/data/train/data.sa.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa3a71ed05fdd3c739d6921be5e01fc005654fcd
--- /dev/null
+++ b/data/train/data.sa.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5353a2f1bfffb9052316153a20aa5c7cf9311e2b723fbe6b5fe365689bd1a66a
+size 1620
diff --git a/data/train/data.sd.json b/data/train/data.sd.json
new file mode 100644
index 0000000000000000000000000000000000000000..7db17da7f5eed4a8af5dfb6727b699fe0fa482b3
--- /dev/null
+++ b/data/train/data.sd.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ae365415111a06b2b79b3dbb2eb20824803fa6d8517aeae0adec35d13b756c6
+size 47727
diff --git a/data/train/data.sh.json b/data/train/data.sh.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf9d307c28ae8a15179016a8b63f36c26605b62f
--- /dev/null
+++ b/data/train/data.sh.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f20f7748942445f54f1f6f5aaebc7c505f8b92b5c7c68f238a0de9a599ffe9b
+size 701559
diff --git a/data/train/data.si.json b/data/train/data.si.json
new file mode 100644
index 0000000000000000000000000000000000000000..5bf169c9b4fcc583637fa66e4ec4ab5c94343601
--- /dev/null
+++ b/data/train/data.si.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:623ef11a0d8c3d5ae8c1c5320a136b15e3f2553dcf1b8c876af0c7ab4b86091d
+size 151561
diff --git a/data/train/data.sk.json b/data/train/data.sk.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7f9aa75328f0f1d553d43b4aa3586ae48b18c7b
--- /dev/null
+++ b/data/train/data.sk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d00b287bc302c7da6e136aab6555a1f05de953846d90d213b727374eac61a44
+size 13384997
diff --git a/data/train/data.sl.json b/data/train/data.sl.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fb8a2ce53ee03b120f45dbdce3d653dee91a0ca
--- /dev/null
+++ b/data/train/data.sl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bac76c4aed49132828aed23f5deee2eb09b05c685afc97770c9b1acd8f33e473
+size 9181878
diff --git a/data/train/data.so.json b/data/train/data.so.json
new file mode 100644
index 0000000000000000000000000000000000000000..42f6d6e68111cdb3a002e7be044f23dbff2b3b7a
--- /dev/null
+++ b/data/train/data.so.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c6c2545a46c68cc5f4356758eb0799581f0d69c3f9eeae642bb37319f8574f0
+size 14996
diff --git a/data/train/data.sq.json b/data/train/data.sq.json
new file mode 100644
index 0000000000000000000000000000000000000000..dfd9dc9037b86fd3d6614a9ee3e94711b440ef9c
--- /dev/null
+++ b/data/train/data.sq.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7851a680414c169047466274768562b7f9e106fa334870a3f5de1f359c590e59
+size 829066
diff --git a/data/train/data.sr.json b/data/train/data.sr.json
new file mode 100644
index 0000000000000000000000000000000000000000..f469031747d64b1d5a3d3648afc1dad818850aa5
--- /dev/null
+++ b/data/train/data.sr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04f72cbcdc9f2b1f9fde305e0dcdf5f86ca7a79e3fbd4883f6489059fabb3c37
+size 8047217
diff --git a/data/train/data.su.json b/data/train/data.su.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c429bcadb29ac640e81e4d16e681009d4ebb7b2
--- /dev/null
+++ b/data/train/data.su.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7d3492b9479d9f3550eac0ba74b87902468df063aded0021011c228acd76d10
+size 19011
diff --git a/data/train/data.sv.json b/data/train/data.sv.json
new file mode 100644
index 0000000000000000000000000000000000000000..bcf00c35d705edad2cb8dbb3774c9aa29805aef1
--- /dev/null
+++ b/data/train/data.sv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69ab1d8679e097233ce29746315cede8f4a3305d370b3b40d2a8e359aa30f143
+size 75208753
diff --git a/data/train/data.sw.json b/data/train/data.sw.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fb01f1c923a772c5fe6a9720716f3fdf2a69972
--- /dev/null
+++ b/data/train/data.sw.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cce9a6a8630ec77f0a2e2903f2c578219058a6473ce19ec44e93a23b4c9ecb39
+size 280459
diff --git a/data/train/data.ta.json b/data/train/data.ta.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca3091ae29e1541bfb316e8f89a7508542446a42
--- /dev/null
+++ b/data/train/data.ta.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0e2d958c236633e4555f02769ad3390f03f7f12d8ba485b7771c830299a27da
+size 948660
diff --git a/data/train/data.te.json b/data/train/data.te.json
new file mode 100644
index 0000000000000000000000000000000000000000..472925577ab64c2af40f6d871d51d738c597c0cf
--- /dev/null
+++ b/data/train/data.te.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:418dc81822ca1b6fdccaf8cc8eff1826f46b746ec9453beabed0ea777b3ba19c
+size 708131
diff --git a/data/train/data.tg.json b/data/train/data.tg.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c70a4dd9c9c23dee4ce2e910d515c8c3834b248
--- /dev/null
+++ b/data/train/data.tg.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3a60b52904fd24df8a2f36727073d9f71ba3ae6078550c6da4a424a4aa011d6
+size 83616
diff --git a/data/train/data.th.json b/data/train/data.th.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c5272d2e2b241f5ff43e35174130893c3cc8e4d
--- /dev/null
+++ b/data/train/data.th.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15168d61f67257a58355fb1d66e5d688b9dc0092a755d7ddfc966f1c983b286c
+size 50233174
diff --git a/data/train/data.tl.json b/data/train/data.tl.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b6557112910e953e9c0f4460ae22d0aeb624fff
--- /dev/null
+++ b/data/train/data.tl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ac401c70a1b8bfc45b835455c2cb4fada99a40b3e03c17c558e09cb36bf8953
+size 1046853
diff --git a/data/train/data.tr.json b/data/train/data.tr.json
new file mode 100644
index 0000000000000000000000000000000000000000..adc7cfe024952b7bc2f7337382eb02d1f255b528
--- /dev/null
+++ b/data/train/data.tr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d83508d761a50e4bdb79a4de2e9ef23dea36401574288b87d9e4bad5606507d
+size 125134131
diff --git a/data/train/data.tt.json b/data/train/data.tt.json
new file mode 100644
index 0000000000000000000000000000000000000000..10686a3767c36e818578e775342df8254f5769e5
--- /dev/null
+++ b/data/train/data.tt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3e64eb410448e3cade5c5866ebe8047d133bef7d0719e61d1aab99230d13b7b
+size 573
diff --git a/data/train/data.ug.json b/data/train/data.ug.json
new file mode 100644
index 0000000000000000000000000000000000000000..44ab2747933e207d73fccef54c302adf3eeaa492
--- /dev/null
+++ b/data/train/data.ug.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:963d19261e87dc86fbad02341c4ecb208d38cb416064ab7cccf9688e1f826130
+size 15572
diff --git a/data/train/data.uk.json b/data/train/data.uk.json
new file mode 100644
index 0000000000000000000000000000000000000000..a0b3283bd87f5da12089e774e555dbd02aedb1f7
--- /dev/null
+++ b/data/train/data.uk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3254b3baf560bf8c0bef1f2f67848b80283eb084e3215a3123d220feb0dd1cf
+size 78505374
diff --git a/data/train/data.ur.json b/data/train/data.ur.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb104787d5baf48c4b0f3037a98b6b1f1e52019d
--- /dev/null
+++ b/data/train/data.ur.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad418badbf7aa6694b50b34317f5a3b05a8ac0dbcda44d112de829a400ad3168
+size 785758
diff --git a/data/train/data.uz.json b/data/train/data.uz.json
new file mode 100644
index 0000000000000000000000000000000000000000..517963c33cd90638f7f72aea51b849bfe49ebd83
--- /dev/null
+++ b/data/train/data.uz.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:124e71a611af31d3e4fcf6546bccd23f76c6720e5db7e0175d69469f0f07c8b6
+size 195207
diff --git a/data/train/data.vep.json b/data/train/data.vep.json
new file mode 100644
index 0000000000000000000000000000000000000000..99366eb388b207786776607d1ce58c0cac9b1970
--- /dev/null
+++ b/data/train/data.vep.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48175c7bd2b729c3b20ebe271b2d1017225422981885fd3cdf4ff53ecc9c91fb
+size 141
diff --git a/data/train/data.vi.json b/data/train/data.vi.json
new file mode 100644
index 0000000000000000000000000000000000000000..d51661249d50614174c1d1acefe8db03ecb555d2
--- /dev/null
+++ b/data/train/data.vi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f832c624863d469814df3613163ce97e8f465e51188ea3e25c2af2d0f0d2900a
+size 140674543
diff --git a/data/train/data.war.json b/data/train/data.war.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b8d2f71a2b0c28b3dacd548c17e40b854c7c61c
--- /dev/null
+++ b/data/train/data.war.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baa830586839a364fd557ddcd8affb9887de75e95f94b8b4a6bcde6a762c59a1
+size 7459
diff --git a/data/train/data.wuu.json b/data/train/data.wuu.json
new file mode 100644
index 0000000000000000000000000000000000000000..8637e4103139e5192ee4a529e9e3b6c36443bef4
--- /dev/null
+++ b/data/train/data.wuu.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4173ddbcc4b9bc0f1e552d0eaa37d0f8a71283f77017b9a15d33e093649cdef8
+size 545
diff --git a/data/train/data.yi.json b/data/train/data.yi.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4eaf737222a35d75de535b2ae862be5d75d5bc9
--- /dev/null
+++ b/data/train/data.yi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2791e2daa11c20632f4d2460aabacfcf6314994df545681865c7cc75e5856dae
+size 51868
diff --git a/data/train/data.yue.json b/data/train/data.yue.json
new file mode 100644
index 0000000000000000000000000000000000000000..d6b1d2f0f31207616f82edce0e409b757af38f44
--- /dev/null
+++ b/data/train/data.yue.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20dba29aa4d9a02082f58c5c65ecf23b843161639dedd74326e915d0635c3ce3
+size 450
diff --git a/data/train/data.zh.json b/data/train/data.zh.json
new file mode 100644
index 0000000000000000000000000000000000000000..52f60191e38a3555b4d7c262425c456b5c259a94
--- /dev/null
+++ b/data/train/data.zh.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36476d48aeac9ffdc4f4e0f5e21d811d93fcbf1b603333aff4e5d631f73324d4
+size 77816471
diff --git a/data/valid/data.af.json b/data/valid/data.af.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.als.json b/data/valid/data.als.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.am.json b/data/valid/data.am.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ar.json b/data/valid/data.ar.json
new file mode 100644
index 0000000000000000000000000000000000000000..369ffe070f7ed60397c9bc680141cb389cb4843a
--- /dev/null
+++ b/data/valid/data.ar.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1459d4873188a1044b4d0a177d8f6c14519d0012aff1163c1bc3059291ec6d24
+size 317304
diff --git a/data/valid/data.arz.json b/data/valid/data.arz.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.as.json b/data/valid/data.as.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.az.json b/data/valid/data.az.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc1256b78f3ce288f55b68d0001d189c052ace30
--- /dev/null
+++ b/data/valid/data.az.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32dd8641533cc9128eb2176fd8787b876e591b61e606373180e0fe4fe27e26f2
+size 28320
diff --git a/data/valid/data.be.json b/data/valid/data.be.json
new file mode 100644
index 0000000000000000000000000000000000000000..66dfb0d4d950822d63850dc5e883d1585daa28ca
--- /dev/null
+++ b/data/valid/data.be.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11fadad246cd248f60a1df70126dce3d3367e7f6d849d415dea8391da8ea87ef
+size 16099
diff --git a/data/valid/data.bg.json b/data/valid/data.bg.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b4a736abe94c53ef4183d2a33b7ec660201828e
--- /dev/null
+++ b/data/valid/data.bg.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd190decfe20b5b1141a54c2ddc0acd48ca52174dfcff17a7ab068413fac7673
+size 309778
diff --git a/data/valid/data.bn.json b/data/valid/data.bn.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c1452e76ee25b2ad5a38624ffbdf01fff358b2b
--- /dev/null
+++ b/data/valid/data.bn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e119ec4ec6b125126581b1a51b34b41d830b97eba90ba0bd726586cf43aa095
+size 78937
diff --git a/data/valid/data.br.json b/data/valid/data.br.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.bs.json b/data/valid/data.bs.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ca.json b/data/valid/data.ca.json
new file mode 100644
index 0000000000000000000000000000000000000000..34efa13a695173729b0988b8360f473b4e3d931e
--- /dev/null
+++ b/data/valid/data.ca.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2e79fb1494523398ac5fe2592f9b5c095e3be5d2ffffcbc1c47d796930b8da6
+size 140812
diff --git a/data/valid/data.ce.json b/data/valid/data.ce.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ceb.json b/data/valid/data.ceb.json
new file mode 100644
index 0000000000000000000000000000000000000000..84f8ffe395bc3f5f997bb495feccab2263856909
--- /dev/null
+++ b/data/valid/data.ceb.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03c8470e547cf7a68d6d017298007a766d8a2d1d6ce4caeb7966971af855f7d0
+size 18136
diff --git a/data/valid/data.ckb.json b/data/valid/data.ckb.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.cs.json b/data/valid/data.cs.json
new file mode 100644
index 0000000000000000000000000000000000000000..7203e95bb62041d59490368ec291a2d190f1befb
--- /dev/null
+++ b/data/valid/data.cs.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bffeed208cc8630be6949fb6df06194b213d95bb50687eb4db05b904616588a
+size 390010
diff --git a/data/valid/data.cy.json b/data/valid/data.cy.json
new file mode 100644
index 0000000000000000000000000000000000000000..147596975b4469265d1ad43139f7c1e77f43b042
--- /dev/null
+++ b/data/valid/data.cy.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5703a912170a6ab26d0def3fbb2a856bec1238df042d86a47aa1593dc832694
+size 13268
diff --git a/data/valid/data.da.json b/data/valid/data.da.json
new file mode 100644
index 0000000000000000000000000000000000000000..893730f923271f62a45f45ca7d6afcdc12f2ca30
--- /dev/null
+++ b/data/valid/data.da.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:773a901d2a6dbf15957f6861007382c20400dc3002feea4aeda8ef6e488a6a83
+size 1116714
diff --git a/data/valid/data.de.json b/data/valid/data.de.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e4d753dd538fa35d6afcc9b044c2044dec32feb
--- /dev/null
+++ b/data/valid/data.de.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb47d8c621654a9860a3917ad1efbe001684f385df2d905547445b8a0851a691
+size 1097357
diff --git a/data/valid/data.dv.json b/data/valid/data.dv.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.el.json b/data/valid/data.el.json
new file mode 100644
index 0000000000000000000000000000000000000000..df087052fca7af2e7155de82a7549b2be518906b
--- /dev/null
+++ b/data/valid/data.el.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28f44d776df870b9016c56f72c500fe0a2cd392cd7061a22028d90e8f018a889
+size 1037026
diff --git a/data/valid/data.eml.json b/data/valid/data.eml.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.en.json b/data/valid/data.en.json
new file mode 100644
index 0000000000000000000000000000000000000000..81c379ea80de269d398821b2d4ce811699a4daa6
--- /dev/null
+++ b/data/valid/data.en.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8444e250ffb8e791696c4e2c188ced6e58cd0804ea480471b0cbd171472d4e5
+size 1100390
diff --git a/data/valid/data.eo.json b/data/valid/data.eo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.es.json b/data/valid/data.es.json
new file mode 100644
index 0000000000000000000000000000000000000000..e71a30e1dc2ae5641d66605b4e6023a99a567907
--- /dev/null
+++ b/data/valid/data.es.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbf3b277056d9b21ff59047335a2125f7308386f6b0b658e81d577b8199af4b1
+size 1121865
diff --git a/data/valid/data.et.json b/data/valid/data.et.json
new file mode 100644
index 0000000000000000000000000000000000000000..9bff87d1c65f187f05e7722d157aea22b4a385ef
--- /dev/null
+++ b/data/valid/data.et.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd40da61a8c186bb7b0cafbd62bcf27cb2b0e0c5c9ce6cf0abf3e1aa5aef28f8
+size 35125
diff --git a/data/valid/data.eu.json b/data/valid/data.eu.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.fa.json b/data/valid/data.fa.json
new file mode 100644
index 0000000000000000000000000000000000000000..d132a332d81d1cc83a09cb1e9931723c6d72cc8d
--- /dev/null
+++ b/data/valid/data.fa.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccd7c261f6fd1b6f71a90ec7b86977bc98118971bece58c46149f17b71bbd6ec
+size 1599035
diff --git a/data/valid/data.fi.json b/data/valid/data.fi.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1a15b4ccac3a65f3ad58185ced215f7e9be61a6
--- /dev/null
+++ b/data/valid/data.fi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3177ddaa59ef467a2adebee0d0e88fd877eb2b55f18193d1a25805dbeb539ade
+size 614814
diff --git a/data/valid/data.fr.json b/data/valid/data.fr.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a59b2ae35babf24c453140c9636c8f3b9a8e852
--- /dev/null
+++ b/data/valid/data.fr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:837483dfb70f8f81562d326cc407ff0e19ccd1aa3a8a05ad2535b12e7b5e5e89
+size 1212141
diff --git a/data/valid/data.fy.json b/data/valid/data.fy.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ga.json b/data/valid/data.ga.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.gd.json b/data/valid/data.gd.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.gl.json b/data/valid/data.gl.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.gu.json b/data/valid/data.gu.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.he.json b/data/valid/data.he.json
new file mode 100644
index 0000000000000000000000000000000000000000..e461ae8753adaf965988e9528a53ebe534193ed1
--- /dev/null
+++ b/data/valid/data.he.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d2766a8ebf5314d5f579bc1d2fee4ea2812e75afda9d3d4041ffedd249627b2
+size 726680
diff --git a/data/valid/data.hi.json b/data/valid/data.hi.json
new file mode 100644
index 0000000000000000000000000000000000000000..297e00b885157d2fe4d28a330bae15633440c71d
--- /dev/null
+++ b/data/valid/data.hi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d82cfb82a5f05b1850f35f149d1cd0637e67b7366841472526dc26a8a2f5c682
+size 167699
diff --git a/data/valid/data.hr.json b/data/valid/data.hr.json
new file mode 100644
index 0000000000000000000000000000000000000000..25804412e1246fae09048caffad8e475c4f240cb
--- /dev/null
+++ b/data/valid/data.hr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2209b385bc301b102228c709bccbe59293b41abd3c4f3d881df2f5399a268321
+size 301363
diff --git a/data/valid/data.ht.json b/data/valid/data.ht.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.hu.json b/data/valid/data.hu.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bd231912452957c6455f90c2d7505ab83c6cb39
--- /dev/null
+++ b/data/valid/data.hu.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eeca604dc0a8998d17179e35ebaa6479b72800c2ee99f8074718fb788977ffb
+size 813071
diff --git a/data/valid/data.hy.json b/data/valid/data.hy.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f8627a70c5394f7b9c5ee0ee00d3c0227c0ecf2
--- /dev/null
+++ b/data/valid/data.hy.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d546fda9f18a2f97de75287773a6f8d7d2a605bb586b0e49bbdaebcbed9f12d
+size 13429
diff --git a/data/valid/data.ia.json b/data/valid/data.ia.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.id.json b/data/valid/data.id.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3f4a68a56d2004a6189e89b0ed881502427c51b
--- /dev/null
+++ b/data/valid/data.id.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5372c0e24734fca223f3bddd2e025e380d1354a67c029d0eaefe50c56933326
+size 642352
diff --git a/data/valid/data.ilo.json b/data/valid/data.ilo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.io.json b/data/valid/data.io.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.is.json b/data/valid/data.is.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac8a72dd2e093c88af53eb29f2eb98df76a26e89
--- /dev/null
+++ b/data/valid/data.is.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a57ef83a80374f08175858c0b7ffa96e14445fb110b2a5ac5b04eb902e063579
+size 42442
diff --git a/data/valid/data.it.json b/data/valid/data.it.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdc784d145d5e6191573ad487b2e05d277a41040
--- /dev/null
+++ b/data/valid/data.it.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8c0e491a4dc4a29cd67ffde465169c34aadd18ad4526f013f29a97b5b77e03d
+size 1236899
diff --git a/data/valid/data.ja.json b/data/valid/data.ja.json
new file mode 100644
index 0000000000000000000000000000000000000000..eaacf852229156c90d27eef9359aa1ebde390998
--- /dev/null
+++ b/data/valid/data.ja.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a884ab5dc4564bdaf12d79642353927d2cae4c73e9909f845a99ced9d42bae0
+size 1026684
diff --git a/data/valid/data.jbo.json b/data/valid/data.jbo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.jv.json b/data/valid/data.jv.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ka.json b/data/valid/data.ka.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3877ccc0baa77369d2e58575fce81cd08d6cb47
--- /dev/null
+++ b/data/valid/data.ka.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6794270d4d04ee8b541046fbac0172b74ff61b8b712281bdedd68c945cdf6251
+size 23082
diff --git a/data/valid/data.kk.json b/data/valid/data.kk.json
new file mode 100644
index 0000000000000000000000000000000000000000..2795cee090db4b6cd5d9b98a94e56838746d82ea
--- /dev/null
+++ b/data/valid/data.kk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c307d5f1701916c6808fb42c25647dbd040bc0bc9755f56a4ed235a577595b66
+size 17139
diff --git a/data/valid/data.km.json b/data/valid/data.km.json
new file mode 100644
index 0000000000000000000000000000000000000000..e638d86fa75222741161d62fb6fd66fb12e4e693
--- /dev/null
+++ b/data/valid/data.km.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b8066025e0261f78c27c9cbf224cd6a9021b1cfd37c3c8bbb602a8dbb26d986
+size 96310
diff --git a/data/valid/data.kn.json b/data/valid/data.kn.json
new file mode 100644
index 0000000000000000000000000000000000000000..dcfd02a06ddc99a789340bb2e54c3ed1986e98ce
--- /dev/null
+++ b/data/valid/data.kn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:470b5a0991d7acd55f5a73c4b56b42e9fe1a94d6f6fce3203cc23ca8a471a485
+size 4842
diff --git a/data/valid/data.ko.json b/data/valid/data.ko.json
new file mode 100644
index 0000000000000000000000000000000000000000..8eb78ef6ebc561f4811c074d5418c25e10728044
--- /dev/null
+++ b/data/valid/data.ko.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da936d2c1b692bed2e7ecc1d730e807e81936a359ecdae4d294becbcd9378ca9
+size 427959
diff --git a/data/valid/data.ku.json b/data/valid/data.ku.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ky.json b/data/valid/data.ky.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.la.json b/data/valid/data.la.json
new file mode 100644
index 0000000000000000000000000000000000000000..75ec0b97fa347d22b5e51eae279f8971c22b9fdc
--- /dev/null
+++ b/data/valid/data.la.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41b96f1c8350be2ce7bc033dacd20f74b9b934ec6825dcd32f1e2e2b9dc051cd
+size 12999
diff --git a/data/valid/data.lb.json b/data/valid/data.lb.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.lmo.json b/data/valid/data.lmo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.lo.json b/data/valid/data.lo.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.lt.json b/data/valid/data.lt.json
new file mode 100644
index 0000000000000000000000000000000000000000..b302a8a3023d2450368c3ae5c268ee00f8af74e2
--- /dev/null
+++ b/data/valid/data.lt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd3006ea811a462c397fdbf011d8581a92469833191d299face8d558648bd996
+size 184045
diff --git a/data/valid/data.lv.json b/data/valid/data.lv.json
new file mode 100644
index 0000000000000000000000000000000000000000..9717d1d7691f6a76e7f97e2be2964aca20223af0
--- /dev/null
+++ b/data/valid/data.lv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ca4f5215becfb9b4048e33c3c7aa86a9fc37a1d8ead4f7c93557e27ddca16e1
+size 103422
diff --git a/data/valid/data.mg.json b/data/valid/data.mg.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.mk.json b/data/valid/data.mk.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbc60abb5eac7fb7e9b5b868689a50f7ab29e122
--- /dev/null
+++ b/data/valid/data.mk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7bc343fda0246ebc94d2f6d3c275aa2472de007ced15eb5b4058eaa898a4d34
+size 35040
diff --git a/data/valid/data.ml.json b/data/valid/data.ml.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.mn.json b/data/valid/data.mn.json
new file mode 100644
index 0000000000000000000000000000000000000000..c00ebd7c5771e07111d685f8fd5fca6953e8c1b4
--- /dev/null
+++ b/data/valid/data.mn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ea929c3e0e761ca955ddd764d19ec94ccc688e8cec2e66dcf9cee40fb9d7dd1
+size 11202
diff --git a/data/valid/data.mr.json b/data/valid/data.mr.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ms.json b/data/valid/data.ms.json
new file mode 100644
index 0000000000000000000000000000000000000000..1928593e1f746a208f8cd44a53a03c0001a5fe8b
--- /dev/null
+++ b/data/valid/data.ms.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8bfecdca3e5428c34ffab6ce8e59b91c6bce44b7e3e747f064a40a6000a60ac
+size 14770
diff --git a/data/valid/data.mt.json b/data/valid/data.mt.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.my.json b/data/valid/data.my.json
new file mode 100644
index 0000000000000000000000000000000000000000..853aa9abe5e0169ee304680001fdc99d7849181d
--- /dev/null
+++ b/data/valid/data.my.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f12471cf4a1d7b137300478f62f78be5ba72ec0b8ac35d443cf754903c7d50e
+size 73510
diff --git a/data/valid/data.nds.json b/data/valid/data.nds.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ne.json b/data/valid/data.ne.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.nl.json b/data/valid/data.nl.json
new file mode 100644
index 0000000000000000000000000000000000000000..abc5204a6d2ba965d37b9f01dfb93f5a85ae86b0
--- /dev/null
+++ b/data/valid/data.nl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b96b129f1a8e7200bfdcb12303c73b6278fb68ed286eee8a3b9a084faca588a
+size 1217537
diff --git a/data/valid/data.nn.json b/data/valid/data.nn.json
new file mode 100644
index 0000000000000000000000000000000000000000..493dcc90899054d258532088cf043ff9e96abd28
--- /dev/null
+++ b/data/valid/data.nn.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea8fe4e641d674f9610064f028a9f22965e55be9dfd99936e9db8ab7d2168cae
+size 6651
diff --git a/data/valid/data.no.json b/data/valid/data.no.json
new file mode 100644
index 0000000000000000000000000000000000000000..94a71e7e9c58387a97f2b2103a7875952c125436
--- /dev/null
+++ b/data/valid/data.no.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdbbedf94bdf4be55668c6d3c768e0cdb5066bef6916b2626c7739450ff878aa
+size 454902
diff --git a/data/valid/data.oc.json b/data/valid/data.oc.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.or.json b/data/valid/data.or.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.pa.json b/data/valid/data.pa.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.pl.json b/data/valid/data.pl.json
new file mode 100644
index 0000000000000000000000000000000000000000..85016e018d2511afd555bb0a4f014934d53b0a67
--- /dev/null
+++ b/data/valid/data.pl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bf73acea9348c8f90b4247e3dbf6c416523a0853f9d1a28572abc7c4f009285
+size 1146853
diff --git a/data/valid/data.pnb.json b/data/valid/data.pnb.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ps.json b/data/valid/data.ps.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.pt.json b/data/valid/data.pt.json
new file mode 100644
index 0000000000000000000000000000000000000000..0982715fc0c8a67cd20cafed6fb20d2191e8b5a7
--- /dev/null
+++ b/data/valid/data.pt.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a595268722e9b6aa29b2c6864103d7d25d08a1a163e5c6e8dcacfe895518f85b
+size 1278217
diff --git a/data/valid/data.rm.json b/data/valid/data.rm.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ro.json b/data/valid/data.ro.json
new file mode 100644
index 0000000000000000000000000000000000000000..0443277c44fbca5cba0aa2237538fd6dd2f0a0ea
--- /dev/null
+++ b/data/valid/data.ro.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16ae9fe1e3b7a9bf4b2a20a500cf9d229ee0c5dae5166658eff96c9d90617663
+size 487787
diff --git a/data/valid/data.ru.json b/data/valid/data.ru.json
new file mode 100644
index 0000000000000000000000000000000000000000..1593e10e3387d8860aa212ac4b75d92f43fa595d
--- /dev/null
+++ b/data/valid/data.ru.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f12b5ffc6f4f43ce34bf69f72259ae592527ae142798eb05c0c0374411ca7ae1
+size 1967775
diff --git a/data/valid/data.sa.json b/data/valid/data.sa.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.sd.json b/data/valid/data.sd.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.sh.json b/data/valid/data.sh.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.si.json b/data/valid/data.si.json
new file mode 100644
index 0000000000000000000000000000000000000000..eff615ea8ef6d6b965bebe8388193bbb43fa0b7d
--- /dev/null
+++ b/data/valid/data.si.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:764a9c1f78e7d184786a9e1c8b3c5edeff9ebaabe9c7d853c1114a39b6a58640
+size 27715
diff --git a/data/valid/data.sk.json b/data/valid/data.sk.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0aeaa77a566c88af4082ee819599e4a838bff8d
--- /dev/null
+++ b/data/valid/data.sk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f93146c5529143f2331001ba4cd91e872242b42db4dea695036f7b98850addfa
+size 396750
diff --git a/data/valid/data.sl.json b/data/valid/data.sl.json
new file mode 100644
index 0000000000000000000000000000000000000000..2baaf14364221169966d5deaead7c9bcc4533969
--- /dev/null
+++ b/data/valid/data.sl.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4652e9dc3270dcb7f8094ad9afa97285afe328dd79115a1c70c5bf9d00cb3b02
+size 261241
diff --git a/data/valid/data.so.json b/data/valid/data.so.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.sq.json b/data/valid/data.sq.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a9d115a9238aefb7d92f30460a4a910f8ab1d31
--- /dev/null
+++ b/data/valid/data.sq.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:365b25ef7e7215a1e9f440f854658c91494badb7dde72aa165f49813a6448f6f
+size 11420
diff --git a/data/valid/data.sr.json b/data/valid/data.sr.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e1cb80b7a2b4bb6a26de741c4e40e9057c8fa8c
--- /dev/null
+++ b/data/valid/data.sr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed1d63b1adf7f91d88b2839066c04dc9265c2874fc3721903b8e25b9ed9e7ac3
+size 366969
diff --git a/data/valid/data.su.json b/data/valid/data.su.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.sv.json b/data/valid/data.sv.json
new file mode 100644
index 0000000000000000000000000000000000000000..c16dc13e27e9e925a7e2c38e54c8e9e25f2e5dad
--- /dev/null
+++ b/data/valid/data.sv.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5ec7278618f2a34f07c47a13bac57d780372e4ae11af0b7fce65ad7c6dc98c6
+size 1029157
diff --git a/data/valid/data.sw.json b/data/valid/data.sw.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ta.json b/data/valid/data.ta.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.te.json b/data/valid/data.te.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.tg.json b/data/valid/data.tg.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.th.json b/data/valid/data.th.json
new file mode 100644
index 0000000000000000000000000000000000000000..eecf61e7894880b8185bbdca1b33f1dc4f04a823
--- /dev/null
+++ b/data/valid/data.th.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7904456ccd0b5b68475d6e687eebb5f3621058d89b1ee61e3d55bd7679526876
+size 328926
diff --git a/data/valid/data.tl.json b/data/valid/data.tl.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.tr.json b/data/valid/data.tr.json
new file mode 100644
index 0000000000000000000000000000000000000000..156ccd1d7b7b8dd1b83c01b0f762038ca19e3673
--- /dev/null
+++ b/data/valid/data.tr.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45f07cb3b7cbc933da91147604e2be58d20b800069409016849e765a19c73ca7
+size 1127712
diff --git a/data/valid/data.tt.json b/data/valid/data.tt.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.ug.json b/data/valid/data.ug.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.uk.json b/data/valid/data.uk.json
new file mode 100644
index 0000000000000000000000000000000000000000..03b75cbf52f37353b4dc5e46573671bbb57f1128
--- /dev/null
+++ b/data/valid/data.uk.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99c672575e0ee4cb444e0e773c3d4280223998d59e4a9cab15079fb9fe5ace5d
+size 109538
diff --git a/data/valid/data.ur.json b/data/valid/data.ur.json
new file mode 100644
index 0000000000000000000000000000000000000000..f90461409394072f896adf2d2339ea4c3db22352
--- /dev/null
+++ b/data/valid/data.ur.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaeb1e3f43460f9e1d2559e4c9c0575f836b26a560ac96c1bd94d43756c40f50
+size 168025
diff --git a/data/valid/data.uz.json b/data/valid/data.uz.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.vep.json b/data/valid/data.vep.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.vi.json b/data/valid/data.vi.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3348e201799d8e01baf95cdd4177debd7f85bd4
--- /dev/null
+++ b/data/valid/data.vi.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d8a1b047d182bc60ca95ad1459039760c29bdda8c19a87a65a661df2a66064d
+size 839188
diff --git a/data/valid/data.war.json b/data/valid/data.war.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.wuu.json b/data/valid/data.wuu.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.yi.json b/data/valid/data.yi.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.yue.json b/data/valid/data.yue.json
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/data/valid/data.zh.json b/data/valid/data.zh.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a51b98751f1c3e49a054628f23604413c347ddc
--- /dev/null
+++ b/data/valid/data.zh.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a1f0581d08cd4c21c40cf00fa8bbc08bd1b1233c35a461a8f8ed2c87abbedf4
+size 993190
diff --git a/dataset_infos.json b/dataset_infos.json
index af63b22458cd7151b566ec84ec66bf00ac2221cd..aff8b9c0af739ddf8788f80da0aa722e35d885f0 100644
--- a/dataset_infos.json
+++ b/dataset_infos.json
@@ -1 +1 @@
-{"en": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3203078691, "num_examples": 6780236, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.en.json": {"num_bytes": 3581637802, "checksum": "32b823a438de3cb8ad21c908d366e8e8b443cdc300665e3dbf5e9aaee015b0af"}}, "download_size": 3581637802, "post_processing_size": null, "dataset_size": 3203078691, "size_in_bytes": 6784716493}, "de": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "de", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 588679009, "num_examples": 1062677, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.de.json": {"num_bytes": 648201529, "checksum": "907d3529fa4dde967e059e69ce8e57a82c7d1dcc45aa88d1145b6ac5fdeeda25"}}, "download_size": 648201529, "post_processing_size": null, "dataset_size": 588679009, "size_in_bytes": 1236880538}, "es": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "es", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 321566270, "num_examples": 625081, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.es.json": {"num_bytes": 356602054, "checksum": "afcfda65b48460774d28222806a05514cec0d36f3a8994753b5caf273f5290fe"}}, "download_size": 356602054, "post_processing_size": null, "dataset_size": 321566270, "size_in_bytes": 678168324}, "fr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 349375263, "num_examples": 642989, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.fr.json": {"num_bytes": 385519610, "checksum": "47d687a72b39ccb9ab4f8ec7225be4ffd642a83090ba763e22936df936f076cf"}}, "download_size": 385519610, "post_processing_size": null, "dataset_size": 349375263, "size_in_bytes": 734894873}, "ru": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 554043247, "num_examples": 871191, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ru.json": {"num_bytes": 602958330, "checksum": "57ee52c4538480f046c4db14b369dec530183d170a33ae5a94eac9b5019f838a"}}, "download_size": 602958330, "post_processing_size": null, "dataset_size": 554043247, "size_in_bytes": 1157001577}, "ja": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ja", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 170280874, "num_examples": 345149, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ja.json": {"num_bytes": 189590993, "checksum": "0ea8e9645ad36e700fa080a001a4bbfdd1dc32fa28aa2bb1717c384a69332670"}}, "download_size": 189590993, "post_processing_size": null, "dataset_size": 170280874, "size_in_bytes": 359871867}, "it": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "it", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 142014290, "num_examples": 304978, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.it.json": {"num_bytes": 159094401, "checksum": "afbe3b387270046f1a53a902be08b1eaa53177d5db392a1272a1371b60b089b9"}}, "download_size": 159094401, "post_processing_size": null, "dataset_size": 142014290, "size_in_bytes": 301108691}, "zh": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 69576524, "num_examples": 162895, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.zh.json": {"num_bytes": 78809663, "checksum": "84d0b5d3fbd9e5003fce7616815161f7f2605eb5a786adaef53dd5c274b4ac9f"}}, "download_size": 78809663, "post_processing_size": null, "dataset_size": 69576524, "size_in_bytes": 148386187}, "pt": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "pt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 112784498, "num_examples": 233191, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.pt.json": {"num_bytes": 125866574, "checksum": "0a3ab7967891ce6dc11b5d00c966a874abd28dbd9bd401b7bc0c5cc71efbf9a9"}}, "download_size": 125866574, "post_processing_size": null, "dataset_size": 112784498, "size_in_bytes": 238651072}, "nl": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 236378057, "num_examples": 452777, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.nl.json": {"num_bytes": 261734229, "checksum": "d878660c32af2b2dbe2e8efc465fd7f083b72b4f93123cb021c85073a1029b91"}}, "download_size": 261734229, "post_processing_size": null, "dataset_size": 236378057, "size_in_bytes": 498112286}, "tr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "tr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 111541042, "num_examples": 266023, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.tr.json": {"num_bytes": 126261845, "checksum": "5e134b8e64a4eb2b6c03fd264dde7238d04fabdf954813d9b70536ab8adfeece"}}, "download_size": 126261845, "post_processing_size": null, "dataset_size": 111541042, "size_in_bytes": 237802887}, "pl": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "pl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 99279579, "num_examples": 198800, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.pl.json": {"num_bytes": 110347602, "checksum": "a6b2cc63c6b82690578b5eec2b94773b8905c58d6de5653c7f72afe400bfd8fe"}}, "download_size": 110347602, "post_processing_size": null, "dataset_size": 99279579, "size_in_bytes": 209627181}, "vi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "vi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 126822231, "num_examples": 263746, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.vi.json": {"num_bytes": 141513733, "checksum": "66d86a95aa223733debe89a4251d890e49e6723d5ae7be3dc5c6789590406999"}}, "download_size": 141513733, "post_processing_size": null, "dataset_size": 126822231, "size_in_bytes": 268335964}, "ar": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ar", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 55248112, "num_examples": 90137, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ar.json": {"num_bytes": 60302171, "checksum": "0be1995ba613ce5ef7527db8ce640152f0b10a7c6aa83eb7ec3a5139fc28792b"}}, "download_size": 60302171, "post_processing_size": null, "dataset_size": 55248112, "size_in_bytes": 115550283}, "id": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "id", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 46374649, "num_examples": 126319, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.id.json": {"num_bytes": 53377583, "checksum": "f881c1bdaa7c3f4d072c7a07ee3af1026c5a0845ba62da85b9119b5231640400"}}, "download_size": 53377583, "post_processing_size": null, "dataset_size": 46374649, "size_in_bytes": 99752232}, "uk": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "uk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 72014565, "num_examples": 117381, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.uk.json": {"num_bytes": 78614914, "checksum": "8070234a23eef0ca0adc849a8e11e383a67509a59403fcd0dcef07e3dc71152e"}}, "download_size": 78614914, "post_processing_size": null, "dataset_size": 72014565, "size_in_bytes": 150629479}, "ro": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ro", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 28873745, "num_examples": 56139, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ro.json": {"num_bytes": 32022363, "checksum": "35db9fbad98780c7cc9903bc9209a2725bdd4348bf3e09a2be1caecbd2fb5422"}}, "download_size": 32022363, "post_processing_size": null, "dataset_size": 28873745, "size_in_bytes": 60896108}, "no": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "no", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27146979, "num_examples": 57595, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.no.json": {"num_bytes": 30369922, "checksum": "c03d2bf1ef411d5d4bea06064049cc16749c10b404a066304d8632be33f3f14b"}}, "download_size": 30369922, "post_processing_size": null, "dataset_size": 27146979, "size_in_bytes": 57516901}, "th": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "th", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 46888113, "num_examples": 66427, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.th.json": {"num_bytes": 50562102, "checksum": "58c224e9d818e49570688c7fa1ee3bc2a7d64c62dff30a6957541dcc7f9c88df"}}, "download_size": 50562102, "post_processing_size": null, "dataset_size": 46888113, "size_in_bytes": 97450215}, "sv": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 67691520, "num_examples": 153523, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.sv.json": {"num_bytes": 76237912, "checksum": "bb89c427bf04d4b18342de86e3ed058ffe5a0f943fb73afcdac374fdbc377a6b"}}, "download_size": 76237912, "post_processing_size": null, "dataset_size": 67691520, "size_in_bytes": 143929432}, "el": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "el", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 39495338, "num_examples": 39482, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.el.json": {"num_bytes": 41745491, "checksum": "6a95e7a41dd207f9b6ad9db0c7662eb87821c2eb6199ab434ba06d002b55ea71"}}, "download_size": 41745491, "post_processing_size": null, "dataset_size": 39495338, "size_in_bytes": 81240829}, "fi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 32912675, "num_examples": 69129, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.fi.json": {"num_bytes": 36755266, "checksum": "c4f74597badc776f55da6b02b18df737a40c46f4625b0e70e28faf56fcaea49b"}}, "download_size": 36755266, "post_processing_size": null, "dataset_size": 32912675, "size_in_bytes": 69667941}, "he": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "he", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 46958548, "num_examples": 78820, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.he.json": {"num_bytes": 51359504, "checksum": "475fea17e03c22339928321167be9717b24f3eb8369d33633bb501739354c4d7"}}, "download_size": 51359504, "post_processing_size": null, "dataset_size": 46958548, "size_in_bytes": 98318052}, "da": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "da", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 48484015, "num_examples": 108306, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.da.json": {"num_bytes": 54519197, "checksum": "571c2bf69513da81b1ca84bb2b21a56b824a145736ec311f070c24a3d20dfcab"}}, "download_size": 54519197, "post_processing_size": null, "dataset_size": 48484015, "size_in_bytes": 103003212}, "cs": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "cs", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 29433325, "num_examples": 60847, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.cs.json": {"num_bytes": 32844002, "checksum": "d64c4586db96e35f64c3411194220221adb93afe1c8aa31c11db0c745141be7e"}}, "download_size": 32844002, "post_processing_size": null, "dataset_size": 29433325, "size_in_bytes": 62277327}, "ko": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ko", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 17819782, "num_examples": 38406, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ko.json": {"num_bytes": 19958758, "checksum": "d9aeb1a433345950312d9f63b2b226f7d59c8e59498958ae224aa738d57667bf"}}, "download_size": 19958758, "post_processing_size": null, "dataset_size": 17819782, "size_in_bytes": 37778540}, "fa": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fa", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 177746834, "num_examples": 315689, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.fa.json": {"num_bytes": 195193088, "checksum": "dc9484e022d71cc836de2f8b348a38fc1290d31f763d20582314d0806e27c912"}}, "download_size": 195193088, "post_processing_size": null, "dataset_size": 177746834, "size_in_bytes": 372939922}, "hi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 42269821, "num_examples": 68627, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.hi.json": {"num_bytes": 46048220, "checksum": "54cb29695d56cf2b66b44697204104356950a746e98f55478525a436752809d4"}}, "download_size": 46048220, "post_processing_size": null, "dataset_size": 42269821, "size_in_bytes": 88318041}, "hu": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 28621034, "num_examples": 50516, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.hu.json": {"num_bytes": 31468795, "checksum": "106931c7b8ade65b282dcf1ecf0fac57532e308cc625300a555eba60ad77f791"}}, "download_size": 31468795, "post_processing_size": null, "dataset_size": 28621034, "size_in_bytes": 60089829}, "sk": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 12371920, "num_examples": 25388, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.sk.json": {"num_bytes": 13781749, "checksum": "8d1e905bec888b97b36f2437c08f18e5618ea5d6d12c0f6b382fb2bf8525a236"}}, "download_size": 13781749, "post_processing_size": null, "dataset_size": 12371920, "size_in_bytes": 26153669}, "lt": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "lt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 8091002, "num_examples": 16339, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.lt.json": {"num_bytes": 9009533, "checksum": "6087756e0020cc80cd246c4b1fe63d37d72196f1f3dec5037cb4b347d8cf8014"}}, "download_size": 9009533, "post_processing_size": null, "dataset_size": 8091002, "size_in_bytes": 17100535}, "et": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "et", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 8286777, "num_examples": 16358, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.et.json": {"num_bytes": 9220886, "checksum": "65e2876f55808eb97a59318e77686fae277b9c23a68c5d62133cef0a32ec9d2f"}}, "download_size": 9220886, "post_processing_size": null, "dataset_size": 8286777, "size_in_bytes": 17507663}, "hr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 9720611, "num_examples": 18650, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.hr.json": {"num_bytes": 10780620, "checksum": "d03e200998e90d5811dd0bf13193023e42b1c25b3d1114014de895e1336d44aa"}}, "download_size": 10780620, "post_processing_size": null, "dataset_size": 9720611, "size_in_bytes": 20501231}, "is": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "is", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1330942, "num_examples": 2473, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.is.json": {"num_bytes": 1471448, "checksum": "eec62be627b967d8a5cdb9de912331f2ecf1af7912accb1bcecea02ca5fcabf2"}}, "download_size": 1471448, "post_processing_size": null, "dataset_size": 1330942, "size_in_bytes": 2802390}, "lv": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "lv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 5047140, "num_examples": 10778, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.lv.json": {"num_bytes": 5648962, "checksum": "18fa458997e0921ec93f9efcdaf33946fe6b5a6ff19bf965a077eec6ddfc272b"}}, "download_size": 5648962, "post_processing_size": null, "dataset_size": 5047140, "size_in_bytes": 10696102}, "ms": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ms", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1551579, "num_examples": 3883, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ms.json": {"num_bytes": 1767684, "checksum": "83aaee6f4e6df670ced098ddc2868e28f96895bae70a73710c1652286ca9230a"}}, "download_size": 1767684, "post_processing_size": null, "dataset_size": 1551579, "size_in_bytes": 3319263}, "bg": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "bg", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 32184900, "num_examples": 39156, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.bg.json": {"num_bytes": 34387573, "checksum": "1040484e6adf757bcb40d987657cad3ff995700a6da13809fce8a3b45ec2aa24"}}, "download_size": 34387573, "post_processing_size": null, "dataset_size": 32184900, "size_in_bytes": 66572473}, "sr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7608481, "num_examples": 14145, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.sr.json": {"num_bytes": 8414188, "checksum": "09c4a9d7a0c486338127fe35fda9ed8fd7206189f162517e55e14ccca4b4784d"}}, "download_size": 8414188, "post_processing_size": null, "dataset_size": 7608481, "size_in_bytes": 16022669}, "ca": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ca", "version": "0.0.0", "splits": {"train": {"name": "train", "num_bytes": 6024971, "num_examples": 7109, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/data.ca.json": {"num_bytes": 6434095, "checksum": "ca1c15a31e109184dedc4367256328cb862d1c9df2408bab6a398b74229bb33e"}}, "download_size": 6434095, "post_processing_size": null, "dataset_size": 6024971, "size_in_bytes": 12459066}}
\ No newline at end of file
+{"en": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3202089080, "num_examples": 6778229, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 989611, "num_examples": 2007, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.en.json": {"num_bytes": 3580537410, "checksum": "bdce76d7893f927534bb649351375dfcb689e97bf8df340266a8455d7b1657a4"}, "data/valid/data.en.json": {"num_bytes": 1100390, "checksum": "f8444e250ffb8e791696c4e2c188ced6e58cd0804ea480471b0cbd171472d4e5"}}, "download_size": 3581637800, "post_processing_size": null, "dataset_size": 3203078691, "size_in_bytes": 6784716491}, "de": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "de", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 587692911, "num_examples": 1060670, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 986114, "num_examples": 2007, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.de.json": {"num_bytes": 647104170, "checksum": "cdf2703481f395d53f33c19e09293366b002954520d4150a3e93381dc2c9bd35"}, "data/valid/data.de.json": {"num_bytes": 1097357, "checksum": "eb47d8c621654a9860a3917ad1efbe001684f385df2d905547445b8a0851a691"}}, "download_size": 648201527, "post_processing_size": null, "dataset_size": 588679025, "size_in_bytes": 1236880552}, "es": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "es", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 320556340, "num_examples": 623059, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1009946, "num_examples": 2022, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.es.json": {"num_bytes": 355480187, "checksum": "e19a787e6653211f0e6237de5aa06abaf75dfa8a60d7ff12dc87f8e9db29a44d"}, "data/valid/data.es.json": {"num_bytes": 1121865, "checksum": "dbf3b277056d9b21ff59047335a2125f7308386f6b0b658e81d577b8199af4b1"}}, "download_size": 356602052, "post_processing_size": null, "dataset_size": 321566286, "size_in_bytes": 678168338}, "fr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 348275004, "num_examples": 640985, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1100275, "num_examples": 2004, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.fr.json": {"num_bytes": 384307467, "checksum": "b0c0a2b8a9848974ace5d0033c2a2c67cc5f9c1846e56b3ded1c808386fe5eee"}, "data/valid/data.fr.json": {"num_bytes": 1212141, "checksum": "837483dfb70f8f81562d326cc407ff0e19ccd1aa3a8a05ad2535b12e7b5e5e89"}}, "download_size": 385519608, "post_processing_size": null, "dataset_size": 349375279, "size_in_bytes": 734894887}, "ru": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 552188755, "num_examples": 869158, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1854492, "num_examples": 2033, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ru.json": {"num_bytes": 600990553, "checksum": "7edaf3ff0b3f94dc78ed01b737c87ea885ad3e18b25f539775f961ea6edc8399"}, "data/valid/data.ru.json": {"num_bytes": 1967775, "checksum": "f12b5ffc6f4f43ce34bf69f72259ae592527ae142798eb05c0c0374411ca7ae1"}}, "download_size": 602958328, "post_processing_size": null, "dataset_size": 554043247, "size_in_bytes": 1157001575}, "ja": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ja", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 169365031, "num_examples": 343147, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 915859, "num_examples": 2002, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ja.json": {"num_bytes": 188564307, "checksum": "cdfea2661848be333ebba4df413aaf133e7d6f95d08e93b7640237026cec0802"}, "data/valid/data.ja.json": {"num_bytes": 1026684, "checksum": "2a884ab5dc4564bdaf12d79642353927d2cae4c73e9909f845a99ced9d42bae0"}}, "download_size": 189590991, "post_processing_size": null, "dataset_size": 170280890, "size_in_bytes": 359871881}, "it": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "it", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 140889289, "num_examples": 302964, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1125017, "num_examples": 2014, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.it.json": {"num_bytes": 157857500, "checksum": "31d5c1d0c1c1a6378ba82430319a038d881cab642383c08088575a5e83ccbf7f"}, "data/valid/data.it.json": {"num_bytes": 1236899, "checksum": "d8c0e491a4dc4a29cd67ffde465169c34aadd18ad4526f013f29a97b5b77e03d"}}, "download_size": 159094399, "post_processing_size": null, "dataset_size": 142014306, "size_in_bytes": 301108705}, "zh": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 68695241, "num_examples": 160893, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 881299, "num_examples": 2002, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.zh.json": {"num_bytes": 77816471, "checksum": "36476d48aeac9ffdc4f4e0f5e21d811d93fcbf1b603333aff4e5d631f73324d4"}, "data/valid/data.zh.json": {"num_bytes": 993190, "checksum": "8a1f0581d08cd4c21c40cf00fa8bbc08bd1b1233c35a461a8f8ed2c87abbedf4"}}, "download_size": 78809661, "post_processing_size": null, "dataset_size": 69576540, "size_in_bytes": 148386201}, "pt": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "pt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 111619119, "num_examples": 231173, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1165395, "num_examples": 2018, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.pt.json": {"num_bytes": 124588355, "checksum": "06aaf895697332f5fe6dc6c4a00c6eeac50b3d0bc8cabe97a5d8adc071a49d81"}, "data/valid/data.pt.json": {"num_bytes": 1278217, "checksum": "a595268722e9b6aa29b2c6864103d7d25d08a1a163e5c6e8dcacfe895518f85b"}}, "download_size": 125866572, "post_processing_size": null, "dataset_size": 112784514, "size_in_bytes": 238651086}, "nl": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 235271716, "num_examples": 450776, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1106357, "num_examples": 2001, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.nl.json": {"num_bytes": 260516690, "checksum": "fa59cab3776009ba9014e40ed0f936454567c4c22942dadae5922ca6285becde"}, "data/valid/data.nl.json": {"num_bytes": 1217537, "checksum": "2b96b129f1a8e7200bfdcb12303c73b6278fb68ed286eee8a3b9a084faca588a"}}, "download_size": 261734227, "post_processing_size": null, "dataset_size": 236378073, "size_in_bytes": 498112300}, "tr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "tr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 110524139, "num_examples": 264008, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1016919, "num_examples": 2015, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.tr.json": {"num_bytes": 125134131, "checksum": "5d83508d761a50e4bdb79a4de2e9ef23dea36401574288b87d9e4bad5606507d"}, "data/valid/data.tr.json": {"num_bytes": 1127712, "checksum": "45f07cb3b7cbc933da91147604e2be58d20b800069409016849e765a19c73ca7"}}, "download_size": 126261843, "post_processing_size": null, "dataset_size": 111541058, "size_in_bytes": 237802901}, "pl": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "pl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 98242932, "num_examples": 196790, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1036663, "num_examples": 2010, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.pl.json": {"num_bytes": 109200747, "checksum": "5b78f6359a4485ef28ebd8f187708c7171f45d70943582aa15b9be43e54694ab"}, "data/valid/data.pl.json": {"num_bytes": 1146853, "checksum": "9bf73acea9348c8f90b4247e3dbf6c416523a0853f9d1a28572abc7c4f009285"}}, "download_size": 110347600, "post_processing_size": null, "dataset_size": 99279595, "size_in_bytes": 209627195}, "vi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "vi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 126047976, "num_examples": 262579, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 774271, "num_examples": 1167, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.vi.json": {"num_bytes": 140674543, "checksum": "f832c624863d469814df3613163ce97e8f465e51188ea3e25c2af2d0f0d2900a"}, "data/valid/data.vi.json": {"num_bytes": 839188, "checksum": "0d8a1b047d182bc60ca95ad1459039760c29bdda8c19a87a65a661df2a66064d"}}, "download_size": 141513731, "post_processing_size": null, "dataset_size": 126822247, "size_in_bytes": 268335978}, "ar": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ar", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 54949705, "num_examples": 89804, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 298407, "num_examples": 333, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ar.json": {"num_bytes": 59984865, "checksum": "e5ede355e54e23a747c3161b8ed34e432fe90d3f419d5486a35d6f22d17f1d48"}, "data/valid/data.ar.json": {"num_bytes": 317304, "checksum": "1459d4873188a1044b4d0a177d8f6c14519d0012aff1163c1bc3059291ec6d24"}}, "download_size": 60302169, "post_processing_size": null, "dataset_size": 55248112, "size_in_bytes": 115550281}, "id": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "id", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 45807886, "num_examples": 124956, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 566779, "num_examples": 1363, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.id.json": {"num_bytes": 52735229, "checksum": "99f189dbd20b85ae225276e50647f82bf503b99c48fd6098b11f47dda8586798"}, "data/valid/data.id.json": {"num_bytes": 642352, "checksum": "a5372c0e24734fca223f3bddd2e025e380d1354a67c029d0eaefe50c56933326"}}, "download_size": 53377581, "post_processing_size": null, "dataset_size": 46374665, "size_in_bytes": 99752246}, "uk": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "uk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 71913409, "num_examples": 117227, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 101172, "num_examples": 154, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.uk.json": {"num_bytes": 78505374, "checksum": "d3254b3baf560bf8c0bef1f2f67848b80283eb084e3215a3123d220feb0dd1cf"}, "data/valid/data.uk.json": {"num_bytes": 109538, "checksum": "99c672575e0ee4cb444e0e773c3d4280223998d59e4a9cab15079fb9fe5ace5d"}}, "download_size": 78614912, "post_processing_size": null, "dataset_size": 72014581, "size_in_bytes": 150629493}, "ro": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ro", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 28430594, "num_examples": 55327, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 443167, "num_examples": 812, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ro.json": {"num_bytes": 31534574, "checksum": "fbed6acbc346664ba882a0633f7d039a55643cdb136401dd70a92ee405308980"}, "data/valid/data.ro.json": {"num_bytes": 487787, "checksum": "16ae9fe1e3b7a9bf4b2a20a500cf9d229ee0c5dae5166658eff96c9d90617663"}}, "download_size": 32022361, "post_processing_size": null, "dataset_size": 28873761, "size_in_bytes": 60896122}, "no": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "no", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26742221, "num_examples": 56679, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 404774, "num_examples": 916, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.no.json": {"num_bytes": 29915018, "checksum": "2a86f2b0d04715915b0883db804a2326dea3a98652ac1ae6f97d4e1e2a476808"}, "data/valid/data.no.json": {"num_bytes": 454902, "checksum": "fdbbedf94bdf4be55668c6d3c768e0cdb5066bef6916b2626c7739450ff878aa"}}, "download_size": 30369920, "post_processing_size": null, "dataset_size": 27146995, "size_in_bytes": 57516915}, "th": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "th", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 46575897, "num_examples": 66135, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 312232, "num_examples": 292, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.th.json": {"num_bytes": 50233174, "checksum": "15168d61f67257a58355fb1d66e5d688b9dc0092a755d7ddfc966f1c983b286c"}, "data/valid/data.th.json": {"num_bytes": 328926, "checksum": "7904456ccd0b5b68475d6e687eebb5f3621058d89b1ee61e3d55bd7679526876"}}, "download_size": 50562100, "post_processing_size": null, "dataset_size": 46888129, "size_in_bytes": 97450229}, "sv": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 66773230, "num_examples": 151508, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 918306, "num_examples": 2015, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.sv.json": {"num_bytes": 75208753, "checksum": "69ab1d8679e097233ce29746315cede8f4a3305d370b3b40d2a8e359aa30f143"}, "data/valid/data.sv.json": {"num_bytes": 1029157, "checksum": "d5ec7278618f2a34f07c47a13bac57d780372e4ae11af0b7fce65ad7c6dc98c6"}}, "download_size": 76237910, "post_processing_size": null, "dataset_size": 67691536, "size_in_bytes": 143929446}, "el": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "el", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 38511192, "num_examples": 38526, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 984162, "num_examples": 956, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.el.json": {"num_bytes": 40708463, "checksum": "78f26bad31241dd2a0d69673354266c54ee3a14c16727574edf8f4643ff4fa70"}, "data/valid/data.el.json": {"num_bytes": 1037026, "checksum": "28f44d776df870b9016c56f72c500fe0a2cd392cd7061a22028d90e8f018a889"}}, "download_size": 41745489, "post_processing_size": null, "dataset_size": 39495354, "size_in_bytes": 81240843}, "fi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 32347167, "num_examples": 68243, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 565524, "num_examples": 886, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.fi.json": {"num_bytes": 36140450, "checksum": "96ea2fdbe15b76784705bf7530cf3a9fe65c4f4cedd7cc415039e9dad2e6a08b"}, "data/valid/data.fi.json": {"num_bytes": 614814, "checksum": "3177ddaa59ef467a2adebee0d0e88fd877eb2b55f18193d1a25805dbeb539ade"}}, "download_size": 36755264, "post_processing_size": null, "dataset_size": 32912691, "size_in_bytes": 69667955}, "he": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "he", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 46285387, "num_examples": 77869, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 673177, "num_examples": 951, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.he.json": {"num_bytes": 50632822, "checksum": "b6bf90449e406791d092f4ad3a11f607722c193f02c3081993bcd7dea72fa559"}, "data/valid/data.he.json": {"num_bytes": 726680, "checksum": "5d2766a8ebf5314d5f579bc1d2fee4ea2812e75afda9d3d4041ffedd249627b2"}}, "download_size": 51359502, "post_processing_size": null, "dataset_size": 46958564, "size_in_bytes": 98318066}, "da": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "da", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 47478594, "num_examples": 106297, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1005437, "num_examples": 2009, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.da.json": {"num_bytes": 53402481, "checksum": "8bbaa638b12f17edf9f89001fa90f5652b4dc64cabf3b22dcd1113ae45f8f837"}, "data/valid/data.da.json": {"num_bytes": 1116714, "checksum": "773a901d2a6dbf15957f6861007382c20400dc3002feea4aeda8ef6e488a6a83"}}, "download_size": 54519195, "post_processing_size": null, "dataset_size": 48484031, "size_in_bytes": 103003226}, "cs": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "cs", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 29081337, "num_examples": 60159, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 352004, "num_examples": 688, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.cs.json": {"num_bytes": 32453990, "checksum": "cbe99fb994e1e7cb22b18d28afa47b1046dc4bc538d9c571d210d90aea80bdda"}, "data/valid/data.cs.json": {"num_bytes": 390010, "checksum": "2bffeed208cc8630be6949fb6df06194b213d95bb50687eb4db05b904616588a"}}, "download_size": 32844000, "post_processing_size": null, "dataset_size": 29433341, "size_in_bytes": 62277341}, "ko": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ko", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 17423874, "num_examples": 37834, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 395924, "num_examples": 572, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ko.json": {"num_bytes": 19530797, "checksum": "c4aabd52cc76758a05828c057c796b31c7b7b63762a4034aa74ce0cfb3b29367"}, "data/valid/data.ko.json": {"num_bytes": 427959, "checksum": "da936d2c1b692bed2e7ecc1d730e807e81936a359ecdae4d294becbcd9378ca9"}}, "download_size": 19958756, "post_processing_size": null, "dataset_size": 17819798, "size_in_bytes": 37778554}, "fa": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "fa", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 176259731, "num_examples": 313682, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 1487119, "num_examples": 2007, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.fa.json": {"num_bytes": 193594051, "checksum": "de8fcd252f7b3e9b52e56b2e24bfc8665c341c8cdfd1c4ae11b5a0a793d17f95"}, "data/valid/data.fa.json": {"num_bytes": 1599035, "checksum": "ccd7c261f6fd1b6f71a90ec7b86977bc98118971bece58c46149f17b71bbd6ec"}}, "download_size": 195193086, "post_processing_size": null, "dataset_size": 177746850, "size_in_bytes": 372939936}, "hi": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 42113447, "num_examples": 68418, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 156390, "num_examples": 209, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.hi.json": {"num_bytes": 45880519, "checksum": "080c6f8a28910a78306a64bb3ff973b8cfce4e789cadb69f0da5d7e96fc84cfd"}, "data/valid/data.hi.json": {"num_bytes": 167699, "checksum": "d82cfb82a5f05b1850f35f149d1cd0637e67b7366841472526dc26a8a2f5c682"}}, "download_size": 46048218, "post_processing_size": null, "dataset_size": 42269837, "size_in_bytes": 88318055}, "hu": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27871546, "num_examples": 49387, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 749488, "num_examples": 1129, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.hu.json": {"num_bytes": 30655722, "checksum": "17509fec8c944f3bc5a20ad3bd385ae011c17d5710b683d4e3be750a0eba2e60"}, "data/valid/data.hu.json": {"num_bytes": 813071, "checksum": "8eeca604dc0a8998d17179e35ebaa6479b72800c2ee99f8074718fb788977ffb"}}, "download_size": 31468793, "post_processing_size": null, "dataset_size": 28621034, "size_in_bytes": 60089827}, "sk": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 12009460, "num_examples": 24763, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 362476, "num_examples": 625, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.sk.json": {"num_bytes": 13384997, "checksum": "7d00b287bc302c7da6e136aab6555a1f05de953846d90d213b727374eac61a44"}, "data/valid/data.sk.json": {"num_bytes": 396750, "checksum": "f93146c5529143f2331001ba4cd91e872242b42db4dea695036f7b98850addfa"}}, "download_size": 13781747, "post_processing_size": null, "dataset_size": 12371936, "size_in_bytes": 26153683}, "lt": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "lt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7924195, "num_examples": 16027, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 166823, "num_examples": 312, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.lt.json": {"num_bytes": 8825486, "checksum": "ed75794218d34e4a3eb854e1bf7e813edf7993312437504909c67f0c26ae1d8c"}, "data/valid/data.lt.json": {"num_bytes": 184045, "checksum": "dd3006ea811a462c397fdbf011d8581a92469833191d299face8d558648bd996"}}, "download_size": 9009531, "post_processing_size": null, "dataset_size": 8091018, "size_in_bytes": 17100549}, "et": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "et", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 8256233, "num_examples": 16274, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 30560, "num_examples": 84, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.et.json": {"num_bytes": 9185759, "checksum": "65872e4d1e738e403c77881096eda2e52e899cbe16b1b030166410d40ba57c9b"}, "data/valid/data.et.json": {"num_bytes": 35125, "checksum": "bd40da61a8c186bb7b0cafbd62bcf27cb2b0e0c5c9ce6cf0abf3e1aa5aef28f8"}}, "download_size": 9220884, "post_processing_size": null, "dataset_size": 8286793, "size_in_bytes": 17507677}, "hr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "hr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 9453241, "num_examples": 18032, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 267386, "num_examples": 618, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.hr.json": {"num_bytes": 10479255, "checksum": "d9756d1c12c57894638ee1025a11406d378b6c6c47a93dbd7be85fb4cc8065fd"}, "data/valid/data.hr.json": {"num_bytes": 301363, "checksum": "2209b385bc301b102228c709bccbe59293b41abd3c4f3d881df2f5399a268321"}}, "download_size": 10780618, "post_processing_size": null, "dataset_size": 9720627, "size_in_bytes": 20501245}, "is": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "is", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1293691, "num_examples": 2379, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 37267, "num_examples": 94, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.is.json": {"num_bytes": 1429004, "checksum": "a3870a71cc60cc616dbcea9d0f80158997d5e625eea63c62378449d55a43c58e"}, "data/valid/data.is.json": {"num_bytes": 42442, "checksum": "a57ef83a80374f08175858c0b7ffa96e14445fb110b2a5ac5b04eb902e063579"}}, "download_size": 1471446, "post_processing_size": null, "dataset_size": 1330958, "size_in_bytes": 2802404}, "lv": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "lv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4955467, "num_examples": 10572, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 91689, "num_examples": 206, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.lv.json": {"num_bytes": 5545538, "checksum": "283c105be7cb93c3276cdde4df055b0a10a6c958c7bbaf730107153a537c57f7"}, "data/valid/data.lv.json": {"num_bytes": 103422, "checksum": "9ca4f5215becfb9b4048e33c3c7aa86a9fc37a1d8ead4f7c93557e27ddca16e1"}}, "download_size": 5648960, "post_processing_size": null, "dataset_size": 5047156, "size_in_bytes": 10696116}, "ms": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ms", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1538651, "num_examples": 3851, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 12944, "num_examples": 32, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ms.json": {"num_bytes": 1752912, "checksum": "8a5de93262a02c1ab436e2c9c07a67efa6bca3369b7281105499636f37bebb8c"}, "data/valid/data.ms.json": {"num_bytes": 14770, "checksum": "d8bfecdca3e5428c34ffab6ce8e59b91c6bce44b7e3e747f064a40a6000a60ac"}}, "download_size": 1767682, "post_processing_size": null, "dataset_size": 1551595, "size_in_bytes": 3319277}, "bg": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "bg", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 31896888, "num_examples": 38759, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 288028, "num_examples": 397, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.bg.json": {"num_bytes": 34077793, "checksum": "5f0186ea6ebdf825c55e2914a614d5ce39af01c15d3f11e8d6e826fbd81b5455"}, "data/valid/data.bg.json": {"num_bytes": 309778, "checksum": "bd190decfe20b5b1141a54c2ddc0acd48ca52174dfcff17a7ab068413fac7673"}}, "download_size": 34387571, "post_processing_size": null, "dataset_size": 32184916, "size_in_bytes": 66572487}, "sr": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "sr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7273319, "num_examples": 13568, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 335178, "num_examples": 577, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.sr.json": {"num_bytes": 8047217, "checksum": "04f72cbcdc9f2b1f9fde305e0dcdf5f86ca7a79e3fbd4883f6489059fabb3c37"}, "data/valid/data.sr.json": {"num_bytes": 366969, "checksum": "ed1d63b1adf7f91d88b2839066c04dc9265c2874fc3721903b8e25b9ed9e7ac3"}}, "download_size": 8414186, "post_processing_size": null, "dataset_size": 7608497, "size_in_bytes": 16022683}, "ca": {"description": "MQA is a multilingual corpus of questions and answers parsed from the Common Crawl. Questions are divided between Frequently Asked Questions (FAQ) pages and Community Question Answering (CQA) pages.", "citation": "\n@misc{debruyn2021mfaq,\n      title={MFAQ: a Multilingual FAQ Dataset}, \n      author={Maxime {De Bruyn} and Ehsan Lotfi and Jeska Buhmann and Walter Daelemans},\n      year={2021},\n      booktitle={MRQA@EMNLP2021},\n}\n", "homepage": "https://huggingface.co/datasets/clips/mqa", "license": "", "features": {"question": {"dtype": "string", "id": null, "_type": "Value"}, "answer": {"dtype": "string", "id": null, "_type": "Value"}, "domain": {"dtype": "string", "id": null, "_type": "Value"}, "domain_index": {"dtype": "int32", "id": null, "_type": "Value"}, "id": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "mfaq_light", "config_name": "ca", "version": "0.0.0", "splits": {"train": {"name": "train", "num_bytes": 5896767, "num_examples": 6882, "dataset_name": "mfaq_light"}, "validation": {"name": "validation", "num_bytes": 128220, "num_examples": 227, "dataset_name": "mfaq_light"}}, "download_checksums": {"data/train/data.ca.json": {"num_bytes": 6293281, "checksum": "c7ecd075efbc5bb08e99a356ccf4942eefba6bc0cdcda477bf9876a6b92c158f"}, "data/valid/data.ca.json": {"num_bytes": 140812, "checksum": "c2e79fb1494523398ac5fe2592f9b5c095e3be5d2ffffcbc1c47d796930b8da6"}}, "download_size": 6434093, "post_processing_size": null, "dataset_size": 6024987, "size_in_bytes": 12459080}}
\ No newline at end of file
diff --git a/mfaq_light.py b/mfaq_light.py
index 48e81e9221d2774ea55e3c9bbf0706b443a5e581..d21a1d680e5fe304544fe218a3630d94a488a0ba 100644
--- a/mfaq_light.py
+++ b/mfaq_light.py
@@ -30,7 +30,8 @@ _CITATION = """
 """
 _VERSION = "0.2"
 _BASE_NAME = ""
-_BASE_URL = "data/data.{}.json"
+_TRAIN_BASE_URL = "data/train/data.{}.json"
+_VALID_BASE_URL = "data/valid/data.{}.json"
 _LANGUAGES = [
     "en", "de", "es", "fr",
     "ru", "ja", "it", "zh", "pt",
@@ -74,15 +75,22 @@ class MFAQLight(datasets.GeneratorBasedBuilder):
         )
     
     def _split_generators(self, dl_manager):
-        filenames = []
+        train_filenames = []
+        valid_filenames = []
         languages = _LANGUAGES if self.config.language == "all" else [self.config.language]
         for language in languages:
-            path = dl_manager.download_and_extract(_BASE_URL.format(language))
-            filenames.append(path)
+            path = dl_manager.download_and_extract(_TRAIN_BASE_URL.format(language))
+            train_filenames.append(path)
+            path = dl_manager.download_and_extract(_VALID_BASE_URL.format(language))
+            valid_filenames.append(path)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
-                gen_kwargs={"filenames": filenames},
+                gen_kwargs={"filenames": train_filenames},
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={"filenames": valid_filenames},
             )
         ]
 
diff --git a/train_valid_split.py b/train_valid_split.py
new file mode 100644
index 0000000000000000000000000000000000000000..eb40e22d9013e8629b595fd68642a565fad08b18
--- /dev/null
+++ b/train_valid_split.py
@@ -0,0 +1,68 @@
+import os
+import json
+import tqdm
+import functools
+import collections
+import multiprocessing
+
+
+def extract_domains(filename):
+    domains = set()
+    with open(filename) as f:
+        for line in f:
+            line = json.loads(line.strip())
+            domains.add(line["domain"])
+    return filename, list(domains)
+
+
+def valid_train_split(filename, mapping=None):
+    previous_domain = ""
+    train = []
+    valid = []
+    domain_data = {"questions": [], "pages": set()}
+    counter = 0
+    with open(filename) as f:
+        for line_txt in f:
+            counter += 1
+            line = json.loads(line_txt.strip())
+            domain = line["domain"]
+            if domain != previous_domain and previous_domain != "":
+                if len(mapping[previous_domain]) > 1:
+                    train.extend(domain_data["questions"])
+                elif len(valid) > 2000:
+                    train.extend(domain_data["questions"])
+                elif len(domain_data["pages"]) > 1:
+                    train.extend(domain_data["questions"])
+                elif len(domain_data["questions"]) < 15:
+                    train.extend(domain_data["questions"])
+                else:
+                    valid.extend(domain_data["questions"])
+                domain_data = {"questions": [], "pages": set()}
+            domain_data["questions"].append(line_txt.strip())
+            domain_data["pages"].add(line["domain_index"])
+            previous_domain = domain
+        train.extend(domain_data["questions"])
+    assert len(train) + len(valid) == counter
+    return train, valid, filename
+
+
+data = [f"data/{e}" for e in os.listdir("data") if e.endswith(".json")]
+with multiprocessing.Pool(os.cpu_count()) as p:
+    domain_count = collections.defaultdict(list)
+    for filename, domains in tqdm.tqdm(p.imap_unordered(extract_domains, data)):
+        language = filename.split(".")[1]
+        for domain in domains:
+            domain_count[domain].append(language)
+
+
+with multiprocessing.Pool(os.cpu_count()) as p:
+    fn = functools.partial(valid_train_split, mapping=domain_count)
+    for train, valid, filename in tqdm.tqdm(p.imap_unordered(fn, data)):
+        train_filename = filename.replace("data/", "data/train/")
+        with open(train_filename, "w+") as f:
+            train = "\n".join(train)
+            f.write(train) 
+        valid_filename = filename.replace("data/", "data/valid/")
+        with open(valid_filename, "w+") as f:
+            valid = "\n".join(valid)
+            f.write(valid)