diff --git a/README.md b/README.md
index c3bebe24982280640b3c75f55fdd9fe7d48ebe6a..4116e78ac4cb00c8822298b42bd3a318dffdeb5d 100644
--- a/README.md
+++ b/README.md
@@ -1,8 +1,10 @@
 ---
-annotations_creators:
-- no-annotation
-language_creators:
-- found
+pretty_name: Opus100
+task_categories:
+- translation
+multilinguality:
+- translation
+task_ids: []
 language:
 - af
 - am
@@ -104,138 +106,21 @@ language:
 - yo
 - zh
 - zu
-license:
-- unknown
-multilinguality:
-- translation
+annotations_creators:
+- no-annotation
+language_creators:
+- found
+source_datasets:
+- extended
 size_categories:
 - 100K<n<1M
 - 10K<n<100K
 - 1K<n<10K
 - 1M<n<10M
 - n<1K
-source_datasets:
-- extended
-task_categories:
-- translation
-task_ids: []
+license:
+- unknown
 paperswithcode_id: opus-100
-pretty_name: OPUS-100
-config_names:
-- af-en
-- am-en
-- an-en
-- ar-de
-- ar-en
-- ar-fr
-- ar-nl
-- ar-ru
-- ar-zh
-- as-en
-- az-en
-- be-en
-- bg-en
-- bn-en
-- br-en
-- bs-en
-- ca-en
-- cs-en
-- cy-en
-- da-en
-- de-en
-- de-fr
-- de-nl
-- de-ru
-- de-zh
-- dz-en
-- el-en
-- en-eo
-- en-es
-- en-et
-- en-eu
-- en-fa
-- en-fi
-- en-fr
-- en-fy
-- en-ga
-- en-gd
-- en-gl
-- en-gu
-- en-ha
-- en-he
-- en-hi
-- en-hr
-- en-hu
-- en-hy
-- en-id
-- en-ig
-- en-is
-- en-it
-- en-ja
-- en-ka
-- en-kk
-- en-km
-- en-kn
-- en-ko
-- en-ku
-- en-ky
-- en-li
-- en-lt
-- en-lv
-- en-mg
-- en-mk
-- en-ml
-- en-mn
-- en-mr
-- en-ms
-- en-mt
-- en-my
-- en-nb
-- en-ne
-- en-nl
-- en-nn
-- en-no
-- en-oc
-- en-or
-- en-pa
-- en-pl
-- en-ps
-- en-pt
-- en-ro
-- en-ru
-- en-rw
-- en-se
-- en-sh
-- en-si
-- en-sk
-- en-sl
-- en-sq
-- en-sr
-- en-sv
-- en-ta
-- en-te
-- en-tg
-- en-th
-- en-tk
-- en-tr
-- en-tt
-- en-ug
-- en-uk
-- en-ur
-- en-uz
-- en-vi
-- en-wa
-- en-xh
-- en-yi
-- en-yo
-- en-zh
-- en-zu
-- fr-nl
-- fr-ru
-- fr-zh
-- nl-ru
-- nl-zh
-- ru-zh
 dataset_info:
 - config_name: af-en
   features:
@@ -247,16 +132,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 135908
+    num_bytes: 135916
     num_examples: 2000
   - name: train
-    num_bytes: 18726247
+    num_bytes: 18726471
     num_examples: 275512
   - name: validation
-    num_bytes: 132769
+    num_bytes: 132777
     num_examples: 2000
-  download_size: 14852797
-  dataset_size: 18994924
+  download_size: 7505036
+  dataset_size: 18995164
 - config_name: am-en
   features:
   - name: translation
@@ -267,16 +152,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 588021
+    num_bytes: 588029
     num_examples: 2000
   - name: train
-    num_bytes: 21950572
+    num_bytes: 21950644
     num_examples: 89027
   - name: validation
-    num_bytes: 566069
+    num_bytes: 566077
     num_examples: 2000
-  download_size: 12630031
-  dataset_size: 23104662
+  download_size: 7004193
+  dataset_size: 23104750
 - config_name: an-en
   features:
   - name: translation
@@ -287,24 +172,10 @@ dataset_info:
         - en
   splits:
   - name: train
-    num_bytes: 438324
+    num_bytes: 438332
     num_examples: 6961
-  download_size: 232976
-  dataset_size: 438324
-- config_name: ar-de
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - ar
-        - de
-  splits:
-  - name: test
-    num_bytes: 238591
-    num_examples: 2000
-  download_size: 161557
-  dataset_size: 238591
+  download_size: 96148
+  dataset_size: 438332
 - config_name: ar-en
   features:
   - name: translation
@@ -315,72 +186,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 331640
+    num_bytes: 331648
     num_examples: 2000
   - name: train
-    num_bytes: 152765684
+    num_bytes: 152766484
     num_examples: 1000000
   - name: validation
-    num_bytes: 2272098
-    num_examples: 2000
-  download_size: 100486814
-  dataset_size: 155369422
-- config_name: ar-fr
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - ar
-        - fr
-  splits:
-  - name: test
-    num_bytes: 547374
-    num_examples: 2000
-  download_size: 334226
-  dataset_size: 547374
-- config_name: ar-nl
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - ar
-        - nl
-  splits:
-  - name: test
-    num_bytes: 212928
-    num_examples: 2000
-  download_size: 144863
-  dataset_size: 212928
-- config_name: ar-ru
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - ar
-        - ru
-  splits:
-  - name: test
-    num_bytes: 808262
-    num_examples: 2000
-  download_size: 441536
-  dataset_size: 808262
-- config_name: ar-zh
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - ar
-        - zh
-  splits:
-  - name: test
-    num_bytes: 713404
+    num_bytes: 2272106
     num_examples: 2000
-  download_size: 438598
-  dataset_size: 713404
+  download_size: 55286865
+  dataset_size: 155370238
 - config_name: as-en
   features:
   - name: translation
@@ -391,16 +206,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 261458
+    num_bytes: 261466
     num_examples: 2000
   - name: train
-    num_bytes: 15634536
+    num_bytes: 15634648
     num_examples: 138479
   - name: validation
-    num_bytes: 248131
+    num_bytes: 248139
     num_examples: 2000
-  download_size: 8794616
-  dataset_size: 16144125
+  download_size: 4183517
+  dataset_size: 16144253
 - config_name: az-en
   features:
   - name: translation
@@ -411,16 +226,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 393101
+    num_bytes: 393109
     num_examples: 2000
   - name: train
-    num_bytes: 56431043
+    num_bytes: 56431259
     num_examples: 262089
   - name: validation
-    num_bytes: 407101
+    num_bytes: 407109
     num_examples: 2000
-  download_size: 34988859
-  dataset_size: 57231245
+  download_size: 18897341
+  dataset_size: 57231477
 - config_name: be-en
   features:
   - name: translation
@@ -431,16 +246,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 166850
+    num_bytes: 166858
     num_examples: 2000
   - name: train
-    num_bytes: 5298444
+    num_bytes: 5298500
     num_examples: 67312
   - name: validation
-    num_bytes: 175197
+    num_bytes: 175205
     num_examples: 2000
-  download_size: 3807669
-  dataset_size: 5640491
+  download_size: 1906088
+  dataset_size: 5640563
 - config_name: bg-en
   features:
   - name: translation
@@ -451,16 +266,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 243743
+    num_bytes: 243751
     num_examples: 2000
   - name: train
-    num_bytes: 108929547
+    num_bytes: 108930347
     num_examples: 1000000
   - name: validation
-    num_bytes: 234840
+    num_bytes: 234848
     num_examples: 2000
-  download_size: 71575310
-  dataset_size: 109408130
+  download_size: 36980744
+  dataset_size: 109408946
 - config_name: bn-en
   features:
   - name: translation
@@ -471,16 +286,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 510093
+    num_bytes: 510101
     num_examples: 2000
   - name: train
-    num_bytes: 249906046
+    num_bytes: 249906846
     num_examples: 1000000
   - name: validation
-    num_bytes: 498406
+    num_bytes: 498414
     num_examples: 2000
-  download_size: 134076596
-  dataset_size: 250914545
+  download_size: 72999655
+  dataset_size: 250915361
 - config_name: br-en
   features:
   - name: translation
@@ -491,16 +306,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 127917
+    num_bytes: 127925
     num_examples: 2000
   - name: train
-    num_bytes: 8538878
+    num_bytes: 8539006
     num_examples: 153447
   - name: validation
-    num_bytes: 133764
+    num_bytes: 133772
     num_examples: 2000
-  download_size: 6881865
-  dataset_size: 8800559
+  download_size: 3323458
+  dataset_size: 8800703
 - config_name: bs-en
   features:
   - name: translation
@@ -511,16 +326,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 168614
+    num_bytes: 168622
     num_examples: 2000
   - name: train
-    num_bytes: 75082148
+    num_bytes: 75082948
     num_examples: 1000000
   - name: validation
-    num_bytes: 172473
+    num_bytes: 172481
     num_examples: 2000
-  download_size: 59514403
-  dataset_size: 75423235
+  download_size: 30746956
+  dataset_size: 75424051
 - config_name: ca-en
   features:
   - name: translation
@@ -531,16 +346,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 205658
+    num_bytes: 205666
     num_examples: 2000
   - name: train
-    num_bytes: 88404710
+    num_bytes: 88405510
     num_examples: 1000000
   - name: validation
-    num_bytes: 212629
+    num_bytes: 212637
     num_examples: 2000
-  download_size: 68438385
-  dataset_size: 88822997
+  download_size: 36267794
+  dataset_size: 88823813
 - config_name: cs-en
   features:
   - name: translation
@@ -551,16 +366,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 205266
+    num_bytes: 205274
     num_examples: 2000
   - name: train
-    num_bytes: 91896919
+    num_bytes: 91897719
     num_examples: 1000000
   - name: validation
-    num_bytes: 219076
+    num_bytes: 219084
     num_examples: 2000
-  download_size: 73028514
-  dataset_size: 92321261
+  download_size: 39673827
+  dataset_size: 92322077
 - config_name: cy-en
   features:
   - name: translation
@@ -571,16 +386,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 124281
+    num_bytes: 124289
     num_examples: 2000
   - name: train
-    num_bytes: 17244748
+    num_bytes: 17244980
     num_examples: 289521
   - name: validation
-    num_bytes: 118848
+    num_bytes: 118856
     num_examples: 2000
-  download_size: 13398765
-  dataset_size: 17487877
+  download_size: 6487005
+  dataset_size: 17488125
 - config_name: da-en
   features:
   - name: translation
@@ -591,16 +406,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 298115
+    num_bytes: 298123
     num_examples: 2000
   - name: train
-    num_bytes: 126424474
+    num_bytes: 126425274
     num_examples: 1000000
   - name: validation
-    num_bytes: 300616
+    num_bytes: 300624
     num_examples: 2000
-  download_size: 91005252
-  dataset_size: 127023205
+  download_size: 50404122
+  dataset_size: 127024021
 - config_name: de-en
   features:
   - name: translation
@@ -611,72 +426,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 330951
+    num_bytes: 330959
     num_examples: 2000
   - name: train
-    num_bytes: 152245956
+    num_bytes: 152246756
     num_examples: 1000000
   - name: validation
-    num_bytes: 332342
-    num_examples: 2000
-  download_size: 116680890
-  dataset_size: 152909249
-- config_name: de-fr
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - de
-        - fr
-  splits:
-  - name: test
-    num_bytes: 458738
-    num_examples: 2000
-  download_size: 311929
-  dataset_size: 458738
-- config_name: de-nl
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - de
-        - nl
-  splits:
-  - name: test
-    num_bytes: 403878
-    num_examples: 2000
-  download_size: 281548
-  dataset_size: 403878
-- config_name: de-ru
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - de
-        - ru
-  splits:
-  - name: test
-    num_bytes: 315771
-    num_examples: 2000
-  download_size: 203225
-  dataset_size: 315771
-- config_name: de-zh
-  features:
-  - name: translation
-    dtype:
-      translation:
-        languages:
-        - de
-        - zh
-  splits:
-  - name: test
-    num_bytes: 280389
+    num_bytes: 332350
     num_examples: 2000
-  download_size: 215301
-  dataset_size: 280389
+  download_size: 67205361
+  dataset_size: 152910065
 - config_name: dz-en
   features:
   - name: translation
@@ -687,10 +446,10 @@ dataset_info:
         - en
   splits:
   - name: train
-    num_bytes: 81154
+    num_bytes: 81162
     num_examples: 624
-  download_size: 37361
-  dataset_size: 81154
+  download_size: 17814
+  dataset_size: 81162
 - config_name: el-en
   features:
   - name: translation
@@ -701,16 +460,16 @@ dataset_info:
         - en
   splits:
   - name: test
-    num_bytes: 302385
+    num_bytes: 302393
     num_examples: 2000
   - name: train
-    num_bytes: 127963903
+    num_bytes: 127964703
     num_examples: 1000000
   - name: validation
-    num_bytes: 291226
+    num_bytes: 291234
     num_examples: 2000
-  download_size: 84137722
-  dataset_size: 128557514
+  download_size: 43973686
+  dataset_size: 128558330
 - config_name: en-eo
   features:
   - name: translation
@@ -721,16 +480,16 @@ dataset_info:
         - eo
   splits:
   - name: test
-    num_bytes: 167378
+    num_bytes: 167386
     num_examples: 2000
   - name: train
-    num_bytes: 24431681
+    num_bytes: 24431953
     num_examples: 337106
   - name: validation
-    num_bytes: 168830
+    num_bytes: 168838
     num_examples: 2000
-  download_size: 19545461
-  dataset_size: 24767889
+  download_size: 9999313
+  dataset_size: 24768177
 - config_name: en-es
   features:
   - name: translation
@@ -741,16 +500,16 @@ dataset_info:
         - es
   splits:
   - name: test
-    num_bytes: 326262
+    num_bytes: 326270
     num_examples: 2000
   - name: train
-    num_bytes: 136643104
+    num_bytes: 136643904
     num_examples: 1000000
   - name: validation
-    num_bytes: 326727
+    num_bytes: 326735
     num_examples: 2000
-  download_size: 100103907
-  dataset_size: 137296093
+  download_size: 55534068
+  dataset_size: 137296909
 - config_name: en-et
   features:
   - name: translation
@@ -761,16 +520,16 @@ dataset_info:
         - et
   splits:
   - name: test
-    num_bytes: 272163
+    num_bytes: 272171
     num_examples: 2000
   - name: train
-    num_bytes: 112298253
+    num_bytes: 112299053
     num_examples: 1000000
   - name: validation
-    num_bytes: 276954
+    num_bytes: 276962
     num_examples: 2000
-  download_size: 83690450
-  dataset_size: 112847370
+  download_size: 46235623
+  dataset_size: 112848186
 - config_name: en-eu
   features:
   - name: translation
@@ -781,16 +540,16 @@ dataset_info:
         - eu
   splits:
   - name: test
-    num_bytes: 280877
+    num_bytes: 280885
     num_examples: 2000
   - name: train
-    num_bytes: 112329285
+    num_bytes: 112330085
     num_examples: 1000000
   - name: validation
-    num_bytes: 281495
+    num_bytes: 281503
     num_examples: 2000
-  download_size: 84805467
-  dataset_size: 112891657
+  download_size: 46389313
+  dataset_size: 112892473
 - config_name: en-fa
   features:
   - name: translation
@@ -801,16 +560,16 @@ dataset_info:
         - fa
   splits:
   - name: test
-    num_bytes: 296548
+    num_bytes: 296556
     num_examples: 2000
   - name: train
-    num_bytes: 125400535
+    num_bytes: 125401335
     num_examples: 1000000
   - name: validation
-    num_bytes: 291121
+    num_bytes: 291129
     num_examples: 2000
-  download_size: 82783248
-  dataset_size: 125988204
+  download_size: 44568447
+  dataset_size: 125989020
 - config_name: en-fi
   features:
   - name: translation
@@ -821,16 +580,16 @@ dataset_info:
         - fi
   splits:
   - name: test
-    num_bytes: 245814
+    num_bytes: 245822
     num_examples: 2000
   - name: train
-    num_bytes: 106024990
+    num_bytes: 106025790
     num_examples: 1000000
   - name: validation
-    num_bytes: 247219
+    num_bytes: 247227
     num_examples: 2000
-  download_size: 79320220
-  dataset_size: 106518023
+  download_size: 42563103
+  dataset_size: 106518839
 - config_name: en-fr
   features:
   - name: translation
@@ -841,16 +600,16 @@ dataset_info:
         - fr
   splits:
   - name: test
-    num_bytes: 469723
+    num_bytes: 469731
     num_examples: 2000
   - name: train
-    num_bytes: 201440450
+    num_bytes: 201441250
     num_examples: 1000000
   - name: validation
-    num_bytes: 481476
+    num_bytes: 481484
     num_examples: 2000
-  download_size: 142251860
-  dataset_size: 202391649
+  download_size: 81009778
+  dataset_size: 202392465
 - config_name: en-fy
   features:
   - name: translation
@@ -861,16 +620,16 @@ dataset_info:
         - fy
   splits:
   - name: test
-    num_bytes: 101238
+    num_bytes: 101246
     num_examples: 2000
   - name: train
-    num_bytes: 3895640
+    num_bytes: 3895688
     num_examples: 54342
   - name: validation
-    num_bytes: 100121
+    num_bytes: 100129
     num_examples: 2000
-  download_size: 2984283
-  dataset_size: 4096999
+  download_size: 1522187
+  dataset_size: 4097063
 - config_name: en-ga
   features:
   - name: translation
@@ -881,16 +640,16 @@ dataset_info:
         - ga
   splits:
   - name: test
-    num_bytes: 503309
+    num_bytes: 503317
     num_examples: 2000
   - name: train
-    num_bytes: 42132510
+    num_bytes: 42132742
     num_examples: 289524
   - name: validation
-    num_bytes: 503209
+    num_bytes: 503217
     num_examples: 2000
-  download_size: 27937448
-  dataset_size: 43139028
+  download_size: 14998873
+  dataset_size: 43139276
 - config_name: en-gd
   features:
   - name: translation
@@ -901,16 +660,16 @@ dataset_info:
         - gd
   splits:
   - name: test
-    num_bytes: 218354
+    num_bytes: 218362
     num_examples: 1606
   - name: train
-    num_bytes: 1254779
+    num_bytes: 1254795
     num_examples: 16316
   - name: validation
-    num_bytes: 203877
+    num_bytes: 203885
     num_examples: 1605
-  download_size: 1124506
-  dataset_size: 1677010
+  download_size: 564053
+  dataset_size: 1677042
 - config_name: en-gl
   features:
   - name: translation
@@ -921,16 +680,16 @@ dataset_info:
         - gl
   splits:
   - name: test
-    num_bytes: 190691
+    num_bytes: 190699
     num_examples: 2000
   - name: train
-    num_bytes: 43327028
+    num_bytes: 43327444
     num_examples: 515344
   - name: validation
-    num_bytes: 193598
+    num_bytes: 193606
     num_examples: 2000
-  download_size: 34084028
-  dataset_size: 43711317
+  download_size: 18056665
+  dataset_size: 43711749
 - config_name: en-gu
   features:
   - name: translation
@@ -941,16 +700,16 @@ dataset_info:
         - gu
   splits:
   - name: test
-    num_bytes: 199725
+    num_bytes: 199733
     num_examples: 2000
   - name: train
-    num_bytes: 33641719
+    num_bytes: 33641975
     num_examples: 318306
   - name: validation
-    num_bytes: 205542
+    num_bytes: 205550
     num_examples: 2000
-  download_size: 19235779
-  dataset_size: 34046986
+  download_size: 9407543
+  dataset_size: 34047258
 - config_name: en-ha
   features:
   - name: translation
@@ -961,16 +720,16 @@ dataset_info:
         - ha
   splits:
   - name: test
-    num_bytes: 407344
+    num_bytes: 407352
     num_examples: 2000
   - name: train
-    num_bytes: 20391884
+    num_bytes: 20391964
     num_examples: 97983
   - name: validation
-    num_bytes: 411518
+    num_bytes: 411526
     num_examples: 2000
-  download_size: 12686187
-  dataset_size: 21210746
+  download_size: 6898482
+  dataset_size: 21210842
 - config_name: en-he
   features:
   - name: translation
@@ -981,16 +740,16 @@ dataset_info:
         - he
   splits:
   - name: test
-    num_bytes: 208467
+    num_bytes: 208475
     num_examples: 2000
   - name: train
-    num_bytes: 91159631
+    num_bytes: 91160431
     num_examples: 1000000
   - name: validation
-    num_bytes: 209438
+    num_bytes: 209446
     num_examples: 2000
-  download_size: 61144758
-  dataset_size: 91577536
+  download_size: 31214136
+  dataset_size: 91578352
 - config_name: en-hi
   features:
   - name: translation
@@ -1001,16 +760,16 @@ dataset_info:
         - hi
   splits:
   - name: test
-    num_bytes: 496570
+    num_bytes: 496578
     num_examples: 2000
   - name: train
-    num_bytes: 124923545
+    num_bytes: 124923977
     num_examples: 534319
   - name: validation
-    num_bytes: 474079
+    num_bytes: 474087
     num_examples: 2000
-  download_size: 65725886
-  dataset_size: 125894194
+  download_size: 35993452
+  dataset_size: 125894642
 - config_name: en-hr
   features:
   - name: translation
@@ -1021,16 +780,16 @@ dataset_info:
         - hr
   splits:
   - name: test
-    num_bytes: 179636
+    num_bytes: 179644
     num_examples: 2000
   - name: train
-    num_bytes: 75309516
+    num_bytes: 75310316
     num_examples: 1000000
   - name: validation
-    num_bytes: 179615
+    num_bytes: 179623
     num_examples: 2000
-  download_size: 59468892
-  dataset_size: 75668767
+  download_size: 30728154
+  dataset_size: 75669583
 - config_name: en-hu
   features:
   - name: translation
@@ -1041,16 +800,16 @@ dataset_info:
         - hu
   splits:
   - name: test
-    num_bytes: 206039
+    num_bytes: 206047
     num_examples: 2000
   - name: train
-    num_bytes: 87483462
+    num_bytes: 87484262
     num_examples: 1000000
   - name: validation
-    num_bytes: 208307
+    num_bytes: 208315
     num_examples: 2000
-  download_size: 67971116
-  dataset_size: 87897808
+  download_size: 35696235
+  dataset_size: 87898624
 - config_name: en-hy
   features:
   - name: translation
@@ -1061,10 +820,10 @@ dataset_info:
         - hy
   splits:
   - name: train
-    num_bytes: 652623
+    num_bytes: 652631
     num_examples: 7059
-  download_size: 422847
-  dataset_size: 652623
+  download_size: 215246
+  dataset_size: 652631
 - config_name: en-id
   features:
   - name: translation
@@ -1075,16 +834,16 @@ dataset_info:
         - id
   splits:
   - name: test
-    num_bytes: 177685
+    num_bytes: 177693
     num_examples: 2000
   - name: train
-    num_bytes: 78698973
+    num_bytes: 78699773
     num_examples: 1000000
   - name: validation
-    num_bytes: 180024
+    num_bytes: 180032
     num_examples: 2000
-  download_size: 57693678
-  dataset_size: 79056682
+  download_size: 29914089
+  dataset_size: 79057498
 - config_name: en-ig
   features:
   - name: translation
@@ -1095,16 +854,16 @@ dataset_info:
         - ig
   splits:
   - name: test
-    num_bytes: 137324
+    num_bytes: 137332
     num_examples: 1843
   - name: train
-    num_bytes: 1612523
+    num_bytes: 1612539
     num_examples: 18415
   - name: validation
-    num_bytes: 135987
+    num_bytes: 135995
     num_examples: 1843
-  download_size: 859440
-  dataset_size: 1885834
+  download_size: 391849
+  dataset_size: 1885866
 - config_name: en-is
   features:
   - name: translation
@@ -1115,16 +874,16 @@ dataset_info:
         - is
   splits:
   - name: test
-    num_bytes: 170879
+    num_bytes: 170887
     num_examples: 2000
   - name: train
-    num_bytes: 73964115
+    num_bytes: 73964915
     num_examples: 1000000
   - name: validation
-    num_bytes: 170632
+    num_bytes: 170640
     num_examples: 2000
-  download_size: 56242149
-  dataset_size: 74305626
+  download_size: 28831218
+  dataset_size: 74306442
 - config_name: en-it
   features:
   - name: translation
@@ -1135,16 +894,16 @@ dataset_info:
         - it
   splits:
   - name: test
-    num_bytes: 299029
+    num_bytes: 299037
     num_examples: 2000
   - name: train
-    num_bytes: 123654286
+    num_bytes: 123655086
     num_examples: 1000000
   - name: validation
-    num_bytes: 294354
+    num_bytes: 294362
     num_examples: 2000
-  download_size: 92133897
-  dataset_size: 124247669
+  download_size: 50903618
+  dataset_size: 124248485
 - config_name: en-ja
   features:
   - name: translation
@@ -1155,16 +914,16 @@ dataset_info:
         - ja
   splits:
   - name: test
-    num_bytes: 190991
+    num_bytes: 190999
     num_examples: 2000
   - name: train
-    num_bytes: 88348569
+    num_bytes: 88349369
     num_examples: 1000000
   - name: validation
-    num_bytes: 191411
+    num_bytes: 191419
     num_examples: 2000
-  download_size: 64817108
-  dataset_size: 88730971
+  download_size: 34452575
+  dataset_size: 88731787
 - config_name: en-ka
   features:
   - name: translation
@@ -1175,16 +934,16 @@ dataset_info:
         - ka
   splits:
   - name: test
-    num_bytes: 256219
+    num_bytes: 256227
     num_examples: 2000
   - name: train
-    num_bytes: 42465402
+    num_bytes: 42465706
     num_examples: 377306
   - name: validation
-    num_bytes: 260408
+    num_bytes: 260416
     num_examples: 2000
-  download_size: 24394633
-  dataset_size: 42982029
+  download_size: 12743188
+  dataset_size: 42982349
 - config_name: en-kk
   features:
   - name: translation
@@ -1195,16 +954,16 @@ dataset_info:
         - kk
   splits:
   - name: test
-    num_bytes: 137656
+    num_bytes: 137664
     num_examples: 2000
   - name: train
-    num_bytes: 7124314
+    num_bytes: 7124378
     num_examples: 79927
   - name: validation
-    num_bytes: 139657
+    num_bytes: 139665
     num_examples: 2000
-  download_size: 4808360
-  dataset_size: 7401627
+  download_size: 2425372
+  dataset_size: 7401707
 - config_name: en-km
   features:
   - name: translation
@@ -1215,56 +974,56 @@ dataset_info:
         - km
   splits:
   - name: test
-    num_bytes: 289019
+    num_bytes: 289027
     num_examples: 2000
   - name: train
-    num_bytes: 19680515
+    num_bytes: 19680611
     num_examples: 111483
   - name: validation
-    num_bytes: 302519
+    num_bytes: 302527
     num_examples: 2000
-  download_size: 10022919
-  dataset_size: 20272053
-- config_name: en-kn
+  download_size: 5193620
+  dataset_size: 20272165
+- config_name: en-ko
   features:
   - name: translation
     dtype:
       translation:
         languages:
         - en
-        - kn
+        - ko
   splits:
   - name: test
-    num_bytes: 77197
-    num_examples: 918
+    num_bytes: 190696
+    num_examples: 2000
   - name: train
-    num_bytes: 1833318
-    num_examples: 14537
+    num_bytes: 93665332
+    num_examples: 1000000
   - name: validation
-    num_bytes: 77599
-    num_examples: 917
-  download_size: 1062554
-  dataset_size: 1988114
-- config_name: en-ko
+    num_bytes: 189368
+    num_examples: 2000
+  download_size: 37602794
+  dataset_size: 94045396
+- config_name: en-kn
   features:
   - name: translation
     dtype:
       translation:
         languages:
         - en
-        - ko
+        - kn
   splits:
   - name: test
-    num_bytes: 190688
-    num_examples: 2000
+    num_bytes: 77205
+    num_examples: 918
   - name: train
-    num_bytes: 93664532
-    num_examples: 1000000
+    num_bytes: 1833334
+    num_examples: 14537
   - name: validation
-    num_bytes: 189360
-    num_examples: 2000
-  download_size: 70383271
-  dataset_size: 94044580
+    num_bytes: 77607
+    num_examples: 917
+  download_size: 525449
+  dataset_size: 1988146
 - config_name: en-ku
   features:
   - name: translation
@@ -1275,16 +1034,16 @@ dataset_info:
         - ku
   splits:
   - name: test
-    num_bytes: 247839
+    num_bytes: 247847
     num_examples: 2000
   - name: train
-    num_bytes: 49107744
+    num_bytes: 49107864
     num_examples: 144844
   - name: validation
-    num_bytes: 239317
+    num_bytes: 239325
     num_examples: 2000
-  download_size: 25358389
-  dataset_size: 49594900
+  download_size: 14252198
+  dataset_size: 49595036
 - config_name: en-ky
   features:
   - name: translation
@@ -1295,16 +1054,16 @@ dataset_info:
         - ky
   splits:
   - name: test
-    num_bytes: 142522
+    num_bytes: 142530
     num_examples: 2000
   - name: train
-    num_bytes: 1879274
+    num_bytes: 1879298
     num_examples: 27215
   - name: validation
-    num_bytes: 138479
+    num_bytes: 138487
     num_examples: 2000
-  download_size: 1338686
-  dataset_size: 2160275
+  download_size: 616902
+  dataset_size: 2160315
 - config_name: en-li
   features:
   - name: translation
@@ -1315,16 +1074,16 @@ dataset_info:
         - li
   splits:
   - name: test
-    num_bytes: 93342
+    num_bytes: 93350
     num_examples: 2000
   - name: train
-    num_bytes: 1628577
+    num_bytes: 1628601
     num_examples: 25535
   - name: validation
-    num_bytes: 92898
+    num_bytes: 92906
     num_examples: 2000
-  download_size: 1040760
-  dataset_size: 1814817
+  download_size: 450092
+  dataset_size: 1814857
 - config_name: en-lt
   features:
   - name: translation
@@ -1335,16 +1094,16 @@ dataset_info:
         - lt
   splits:
   - name: test
-    num_bytes: 482607
+    num_bytes: 482615
     num_examples: 2000
   - name: train
-    num_bytes: 177060244
+    num_bytes: 177061044
     num_examples: 1000000
   - name: validation
-    num_bytes: 469109
+    num_bytes: 469117
     num_examples: 2000
-  download_size: 124444053
-  dataset_size: 178011960
+  download_size: 69388131
+  dataset_size: 178012776
 - config_name: en-lv
   features:
   - name: translation
@@ -1355,16 +1114,16 @@ dataset_info:
         - lv
   splits:
   - name: test
-    num_bytes: 536568
+    num_bytes: 536576
     num_examples: 2000
   - name: train
-    num_bytes: 206051049
+    num_bytes: 206051849
     num_examples: 1000000
   - name: validation
-    num_bytes: 522064
+    num_bytes: 522072
     num_examples: 2000
-  download_size: 140538527
-  dataset_size: 207109681
+  download_size: 78952903
+  dataset_size: 207110497
 - config_name: en-mg
   features:
   - name: translation
@@ -1375,16 +1134,16 @@ dataset_info:
         - mg
   splits:
   - name: test
-    num_bytes: 525059
+    num_bytes: 525067
     num_examples: 2000
   - name: train
-    num_bytes: 130865169
+    num_bytes: 130865649
     num_examples: 590771
   - name: validation
-    num_bytes: 511163
+    num_bytes: 511171
     num_examples: 2000
-  download_size: 91102165
-  dataset_size: 131901391
+  download_size: 52470504
+  dataset_size: 131901887
 - config_name: en-mk
   features:
   - name: translation
@@ -1395,16 +1154,16 @@ dataset_info:
         - mk
   splits:
   - name: test
-    num_bytes: 308926
+    num_bytes: 308934
     num_examples: 2000
   - name: train
-    num_bytes: 117068689
+    num_bytes: 117069489
     num_examples: 1000000
   - name: validation
-    num_bytes: 305490
+    num_bytes: 305498
     num_examples: 2000
-  download_size: 76810811
-  dataset_size: 117683105
+  download_size: 39517761
+  dataset_size: 117683921
 - config_name: en-ml
   features:
   - name: translation
@@ -1415,16 +1174,16 @@ dataset_info:
         - ml
   splits:
   - name: test
-    num_bytes: 340618
+    num_bytes: 340626
     num_examples: 2000
   - name: train
-    num_bytes: 199971079
+    num_bytes: 199971743
     num_examples: 822746
   - name: validation
-    num_bytes: 334451
+    num_bytes: 334459
     num_examples: 2000
-  download_size: 95497482
-  dataset_size: 200646148
+  download_size: 48654808
+  dataset_size: 200646828
 - config_name: en-mn
   features:
   - name: translation
@@ -1435,10 +1194,10 @@ dataset_info:
         - mn
   splits:
   - name: train
-    num_bytes: 250770
+    num_bytes: 250778
     num_examples: 4294
-  download_size: 85037
-  dataset_size: 250770
+  download_size: 42039
+  dataset_size: 250778
 - config_name: en-mr
   features:
   - name: translation
@@ -1449,16 +1208,16 @@ dataset_info:
         - mr
   splits:
   - name: test
-    num_bytes: 238604
+    num_bytes: 238612
     num_examples: 2000
   - name: train
-    num_bytes: 2724107
+    num_bytes: 2724131
     num_examples: 27007
   - name: validation
-    num_bytes: 235532
+    num_bytes: 235540
     num_examples: 2000
-  download_size: 1838618
-  dataset_size: 3198243
+  download_size: 910211
+  dataset_size: 3198283
 - config_name: en-ms
   features:
   - name: translation
@@ -1469,16 +1228,16 @@ dataset_info:
         - ms
   splits:
   - name: test
-    num_bytes: 179697
+    num_bytes: 179705
     num_examples: 2000
   - name: train
-    num_bytes: 76828845
+    num_bytes: 76829645
     num_examples: 1000000
   - name: validation
-    num_bytes: 180175
+    num_bytes: 180183
     num_examples: 2000
-  download_size: 57412836
-  dataset_size: 77188717
+  download_size: 29807607
+  dataset_size: 77189533
 - config_name: en-mt
   features:
   - name: translation
@@ -1489,16 +1248,16 @@ dataset_info:
         - mt
   splits:
   - name: test
-    num_bytes: 566126
+    num_bytes: 566134
     num_examples: 2000
   - name: train
-    num_bytes: 222221596
+    num_bytes: 222222396
     num_examples: 1000000
   - name: validation
-    num_bytes: 594378
+    num_bytes: 594386
     num_examples: 2000
-  download_size: 147836637
-  dataset_size: 223382100
+  download_size: 84757608
+  dataset_size: 223382916
 - config_name: en-my
   features:
   - name: translation
@@ -1509,16 +1268,16 @@ dataset_info:
         - my
   splits:
   - name: test
-    num_bytes: 337343
+    num_bytes: 337351
     num_examples: 2000
   - name: train
-    num_bytes: 3673477
+    num_bytes: 3673501
     num_examples: 24594
   - name: validation
-    num_bytes: 336147
+    num_bytes: 336155
     num_examples: 2000
-  download_size: 1952573
-  dataset_size: 4346967
+  download_size: 1038600
+  dataset_size: 4347007
 - config_name: en-nb
   features:
   - name: translation
@@ -1529,16 +1288,16 @@ dataset_info:
         - nb
   splits:
   - name: test
-    num_bytes: 334109
+    num_bytes: 334117
     num_examples: 2000
   - name: train
-    num_bytes: 13611589
+    num_bytes: 13611709
     num_examples: 142906
   - name: validation
-    num_bytes: 324392
+    num_bytes: 324400
     num_examples: 2000
-  download_size: 10630769
-  dataset_size: 14270090
+  download_size: 5706626
+  dataset_size: 14270226
 - config_name: en-ne
   features:
   - name: translation
@@ -1549,16 +1308,16 @@ dataset_info:
         - ne
   splits:
   - name: test
-    num_bytes: 186519
+    num_bytes: 186527
     num_examples: 2000
   - name: train
-    num_bytes: 44135952
+    num_bytes: 44136280
     num_examples: 406381
   - name: validation
-    num_bytes: 204912
+    num_bytes: 204920
     num_examples: 2000
-  download_size: 24107523
-  dataset_size: 44527383
+  download_size: 11711988
+  dataset_size: 44527727
 - config_name: en-nl
   features:
   - name: translation
@@ -1569,16 +1328,16 @@ dataset_info:
         - nl
   splits:
   - name: test
-    num_bytes: 282747
+    num_bytes: 282755
     num_examples: 2000
   - name: train
-    num_bytes: 112326273
+    num_bytes: 112327073
     num_examples: 1000000
   - name: validation
-    num_bytes: 270932
+    num_bytes: 270940
     num_examples: 2000
-  download_size: 82923916
-  dataset_size: 112879952
+  download_size: 45374708
+  dataset_size: 112880768
 - config_name: en-nn
   features:
   - name: translation
@@ -1589,16 +1348,16 @@ dataset_info:
         - nn
   splits:
   - name: test
-    num_bytes: 178999
+    num_bytes: 179007
     num_examples: 2000
   - name: train
-    num_bytes: 32924429
+    num_bytes: 32924821
     num_examples: 486055
   - name: validation
-    num_bytes: 187642
+    num_bytes: 187650
     num_examples: 2000
-  download_size: 25184676
-  dataset_size: 33291070
+  download_size: 12742134
+  dataset_size: 33291478
 - config_name: en-no
   features:
   - name: translation
@@ -1609,16 +1368,16 @@ dataset_info:
         - 'no'
   splits:
   - name: test
-    num_bytes: 173320
+    num_bytes: 173328
     num_examples: 2000
   - name: train
-    num_bytes: 74105483
+    num_bytes: 74106283
     num_examples: 1000000
   - name: validation
-    num_bytes: 178005
+    num_bytes: 178013
     num_examples: 2000
-  download_size: 56277000
-  dataset_size: 74456808
+  download_size: 28851262
+  dataset_size: 74457624
 - config_name: en-oc
   features:
   - name: translation
@@ -1629,16 +1388,16 @@ dataset_info:
         - oc
   splits:
   - name: test
-    num_bytes: 82342
+    num_bytes: 82350
     num_examples: 2000
   - name: train
-    num_bytes: 1627174
+    num_bytes: 1627206
     num_examples: 35791
   - name: validation
-    num_bytes: 81642
+    num_bytes: 81650
     num_examples: 2000
-  download_size: 1308338
-  dataset_size: 1791158
+  download_size: 607192
+  dataset_size: 1791206
 - config_name: en-or
   features:
   - name: translation
@@ -1649,16 +1408,16 @@ dataset_info:
         - or
   splits:
   - name: test
-    num_bytes: 163939
+    num_bytes: 163947
     num_examples: 1318
   - name: train
-    num_bytes: 1500733
+    num_bytes: 1500749
     num_examples: 14273
   - name: validation
-    num_bytes: 155323
+    num_bytes: 155331
     num_examples: 1317
-  download_size: 1019971
-  dataset_size: 1819995
+  download_size: 499401
+  dataset_size: 1820027
 - config_name: en-pa
   features:
   - name: translation
@@ -1669,16 +1428,16 @@ dataset_info:
         - pa
   splits:
   - name: test
-    num_bytes: 133901
+    num_bytes: 133909
     num_examples: 2000
   - name: train
-    num_bytes: 8509140
+    num_bytes: 8509228
     num_examples: 107296
   - name: validation
-    num_bytes: 136188
+    num_bytes: 136196
     num_examples: 2000
-  download_size: 5315298
-  dataset_size: 8779229
+  download_size: 2589682
+  dataset_size: 8779333
 - config_name: en-pl
   features:
   - name: translation
@@ -1689,16 +1448,16 @@ dataset_info:
         - pl
   splits:
   - name: test
-    num_bytes: 212495
+    num_bytes: 212503
     num_examples: 2000
   - name: train
-    num_bytes: 95247723
+    num_bytes: 95248523
     num_examples: 1000000
   - name: validation
-    num_bytes: 218208
+    num_bytes: 218216
     num_examples: 2000
-  download_size: 73574044
-  dataset_size: 95678426
+  download_size: 39320454
+  dataset_size: 95679242
 - config_name: en-ps
   features:
   - name: translation
@@ -1709,16 +1468,16 @@ dataset_info:
         - ps
   splits:
   - name: test
-    num_bytes: 92995
+    num_bytes: 93003
     num_examples: 2000
   - name: train
-    num_bytes: 4436512
+    num_bytes: 4436576
     num_examples: 79127
   - name: validation
-    num_bytes: 95156
+    num_bytes: 95164
     num_examples: 2000
-  download_size: 2851899
-  dataset_size: 4624663
+  download_size: 1223087
+  dataset_size: 4624743
 - config_name: en-pt
   features:
   - name: translation
@@ -1729,16 +1488,16 @@ dataset_info:
         - pt
   splits:
   - name: test
-    num_bytes: 296114
+    num_bytes: 296122
     num_examples: 2000
   - name: train
-    num_bytes: 118242849
+    num_bytes: 118243649
     num_examples: 1000000
   - name: validation
-    num_bytes: 292074
+    num_bytes: 292082
     num_examples: 2000
-  download_size: 87661907
-  dataset_size: 118831037
+  download_size: 48087550
+  dataset_size: 118831853
 - config_name: en-ro
   features:
   - name: translation
@@ -1749,16 +1508,16 @@ dataset_info:
         - ro
   splits:
   - name: test
-    num_bytes: 198639
+    num_bytes: 198647
     num_examples: 2000
   - name: train
-    num_bytes: 85249051
+    num_bytes: 85249851
     num_examples: 1000000
   - name: validation
-    num_bytes: 199164
+    num_bytes: 199172
     num_examples: 2000
-  download_size: 66294317
-  dataset_size: 85646854
+  download_size: 35032743
+  dataset_size: 85647670
 - config_name: en-ru
   features:
   - name: translation
@@ -1769,16 +1528,16 @@ dataset_info:
         - ru
   splits:
   - name: test
-    num_bytes: 490976
+    num_bytes: 490984
     num_examples: 2000
   - name: train
-    num_bytes: 195100937
+    num_bytes: 195101737
     num_examples: 1000000
   - name: validation
-    num_bytes: 490238
+    num_bytes: 490246
     num_examples: 2000
-  download_size: 124460816
-  dataset_size: 196082151
+  download_size: 68501634
+  dataset_size: 196082967
 - config_name: en-rw
   features:
   - name: translation
@@ -1789,16 +1548,16 @@ dataset_info:
         - rw
   splits:
   - name: test
-    num_bytes: 136189
+    num_bytes: 136197
     num_examples: 2000
   - name: train
-    num_bytes: 15286159
+    num_bytes: 15286303
     num_examples: 173823
   - name: validation
-    num_bytes: 134957
+    num_bytes: 134965
     num_examples: 2000
-  download_size: 10093708
-  dataset_size: 15557305
+  download_size: 5233241
+  dataset_size: 15557465
 - config_name: en-se
   features:
   - name: translation
@@ -1809,16 +1568,16 @@ dataset_info:
         - se
   splits:
   - name: test
-    num_bytes: 85697
+    num_bytes: 85705
     num_examples: 2000
   - name: train
-    num_bytes: 2047380
+    num_bytes: 2047412
     num_examples: 35907
   - name: validation
-    num_bytes: 83664
+    num_bytes: 83672
     num_examples: 2000
-  download_size: 1662845
-  dataset_size: 2216741
+  download_size: 806982
+  dataset_size: 2216789
 - config_name: en-sh
   features:
   - name: translation
@@ -1829,16 +1588,16 @@ dataset_info:
         - sh
   splits:
   - name: test
-    num_bytes: 569479
+    num_bytes: 569487
     num_examples: 2000
   - name: train
-    num_bytes: 60900023
+    num_bytes: 60900239
     num_examples: 267211
   - name: validation
-    num_bytes: 555594
+    num_bytes: 555602
     num_examples: 2000
-  download_size: 39988454
-  dataset_size: 62025096
+  download_size: 22357505
+  dataset_size: 62025328
 - config_name: en-si
   features:
   - name: translation
@@ -1849,16 +1608,16 @@ dataset_info:
         - si
   splits:
   - name: test
-    num_bytes: 271735
+    num_bytes: 271743
     num_examples: 2000
   - name: train
-    num_bytes: 114950891
+    num_bytes: 114951675
     num_examples: 979109
   - name: validation
-    num_bytes: 271236
+    num_bytes: 271244
     num_examples: 2000
-  download_size: 66124160
-  dataset_size: 115493862
+  download_size: 33247484
+  dataset_size: 115494662
 - config_name: en-sk
   features:
   - name: translation
@@ -1869,16 +1628,16 @@ dataset_info:
         - sk
   splits:
   - name: test
-    num_bytes: 258034
+    num_bytes: 258042
     num_examples: 2000
   - name: train
-    num_bytes: 111743068
+    num_bytes: 111743868
     num_examples: 1000000
   - name: validation
-    num_bytes: 255462
+    num_bytes: 255470
     num_examples: 2000
-  download_size: 85223330
-  dataset_size: 112256564
+  download_size: 46618395
+  dataset_size: 112257380
 - config_name: en-sl
   features:
   - name: translation
@@ -1889,16 +1648,16 @@ dataset_info:
         - sl
   splits:
   - name: test
-    num_bytes: 205470
+    num_bytes: 205478
     num_examples: 2000
   - name: train
-    num_bytes: 90270157
+    num_bytes: 90270957
     num_examples: 1000000
   - name: validation
-    num_bytes: 198654
+    num_bytes: 198662
     num_examples: 2000
-  download_size: 70708189
-  dataset_size: 90674281
+  download_size: 37536724
+  dataset_size: 90675097
 - config_name: en-sq
   features:
   - name: translation
@@ -1909,16 +1668,16 @@ dataset_info:
         - sq
   splits:
   - name: test
-    num_bytes: 275371
+    num_bytes: 275379
     num_examples: 2000
   - name: train
-    num_bytes: 105745181
+    num_bytes: 105745981
     num_examples: 1000000
   - name: validation
-    num_bytes: 267304
+    num_bytes: 267312
     num_examples: 2000
-  download_size: 78817895
-  dataset_size: 106287856
+  download_size: 42697338
+  dataset_size: 106288672
 - config_name: en-sr
   features:
   - name: translation
@@ -1929,16 +1688,16 @@ dataset_info:
         - sr
   splits:
   - name: test
-    num_bytes: 180224
+    num_bytes: 180232
     num_examples: 2000
   - name: train
-    num_bytes: 75726035
+    num_bytes: 75726835
     num_examples: 1000000
   - name: validation
-    num_bytes: 184238
+    num_bytes: 184246
     num_examples: 2000
-  download_size: 60263688
-  dataset_size: 76090497
+  download_size: 31260575
+  dataset_size: 76091313
 - config_name: en-sv
   features:
   - name: translation
@@ -1949,16 +1708,16 @@ dataset_info:
         - sv
   splits:
   - name: test
-    num_bytes: 271006
+    num_bytes: 271014
     num_examples: 2000
   - name: train
-    num_bytes: 116985153
+    num_bytes: 116985953
     num_examples: 1000000
   - name: validation
-    num_bytes: 279986
+    num_bytes: 279994
     num_examples: 2000
-  download_size: 85032127
-  dataset_size: 117536145
+  download_size: 46694960
+  dataset_size: 117536961
 - config_name: en-ta
   features:
   - name: translation
@@ -1969,16 +1728,16 @@ dataset_info:
         - ta
   splits:
   - name: test
-    num_bytes: 351982
+    num_bytes: 351990
     num_examples: 2000
   - name: train
-    num_bytes: 74044340
+    num_bytes: 74044524
     num_examples: 227014
   - name: validation
-    num_bytes: 335549
+    num_bytes: 335557
     num_examples: 2000
-  download_size: 33642694
-  dataset_size: 74731871
+  download_size: 17652443
+  dataset_size: 74732071
 - config_name: en-te
   features:
   - name: translation
@@ -1989,16 +1748,16 @@ dataset_info:
         - te
   splits:
   - name: test
-    num_bytes: 190587
+    num_bytes: 190595
     num_examples: 2000
   - name: train
-    num_bytes: 6688569
+    num_bytes: 6688625
     num_examples: 64352
   - name: validation
-    num_bytes: 193658
+    num_bytes: 193666
     num_examples: 2000
-  download_size: 4047667
-  dataset_size: 7072814
+  download_size: 2011832
+  dataset_size: 7072886
 - config_name: en-tg
   features:
   - name: translation
@@ -2009,16 +1768,16 @@ dataset_info:
         - tg
   splits:
   - name: test
-    num_bytes: 372112
+    num_bytes: 372120
     num_examples: 2000
   - name: train
-    num_bytes: 35477017
+    num_bytes: 35477177
     num_examples: 193882
   - name: validation
-    num_bytes: 371720
+    num_bytes: 371728
     num_examples: 2000
-  download_size: 21242668
-  dataset_size: 36220849
+  download_size: 11389877
+  dataset_size: 36221025
 - config_name: en-th
   features:
   - name: translation
@@ -2029,16 +1788,16 @@ dataset_info:
         - th
   splits:
   - name: test
-    num_bytes: 290573
+    num_bytes: 290581
     num_examples: 2000
   - name: train
-    num_bytes: 132820231
+    num_bytes: 132821031
     num_examples: 1000000
   - name: validation
-    num_bytes: 288358
+    num_bytes: 288366
     num_examples: 2000
-  download_size: 75539987
-  dataset_size: 133399162
+  download_size: 38147204
+  dataset_size: 133399978
 - config_name: en-tk
   features:
   - name: translation
@@ -2049,16 +1808,16 @@ dataset_info:
         - tk
   splits:
   - name: test
-    num_bytes: 83878
+    num_bytes: 83886
     num_examples: 1852
   - name: train
-    num_bytes: 719617
+    num_bytes: 719633
     num_examples: 13110
   - name: validation
-    num_bytes: 81006
+    num_bytes: 81014
     num_examples: 1852
-  download_size: 417756
-  dataset_size: 884501
+  download_size: 157481
+  dataset_size: 884533
 - config_name: en-tr
   features:
   - name: translation
@@ -2069,16 +1828,16 @@ dataset_info:
         - tr
   splits:
   - name: test
-    num_bytes: 183825
+    num_bytes: 183833
     num_examples: 2000
   - name: train
-    num_bytes: 78945565
+    num_bytes: 78946365
     num_examples: 1000000
   - name: validation
-    num_bytes: 181909
+    num_bytes: 181917
     num_examples: 2000
-  download_size: 60364921
-  dataset_size: 79311299
+  download_size: 30892429
+  dataset_size: 79312115
 - config_name: en-tt
   features:
   - name: translation
@@ -2089,16 +1848,16 @@ dataset_info:
         - tt
   splits:
   - name: test
-    num_bytes: 693268
+    num_bytes: 693276
     num_examples: 2000
   - name: train
-    num_bytes: 35313170
+    num_bytes: 35313258
     num_examples: 100843
   - name: validation
-    num_bytes: 701662
+    num_bytes: 701670
     num_examples: 2000
-  download_size: 18786998
-  dataset_size: 36708100
+  download_size: 9940523
+  dataset_size: 36708204
 - config_name: en-ug
   features:
   - name: translation
@@ -2109,16 +1868,16 @@ dataset_info:
         - ug
   splits:
   - name: test
-    num_bytes: 620873
+    num_bytes: 620881
     num_examples: 2000
   - name: train
-    num_bytes: 31576516
+    num_bytes: 31576580
     num_examples: 72170
   - name: validation
-    num_bytes: 631228
+    num_bytes: 631236
     num_examples: 2000
-  download_size: 16011372
-  dataset_size: 32828617
+  download_size: 8687743
+  dataset_size: 32828697
 - config_name: en-uk
   features:
   - name: translation
@@ -2129,16 +1888,16 @@ dataset_info:
         - uk
   splits:
   - name: test
-    num_bytes: 249742
+    num_bytes: 249750
     num_examples: 2000
   - name: train
-    num_bytes: 104229556
+    num_bytes: 104230356
     num_examples: 1000000
   - name: validation
-    num_bytes: 247123
+    num_bytes: 247131
     num_examples: 2000
-  download_size: 71155682
-  dataset_size: 104726421
+  download_size: 37415496
+  dataset_size: 104727237
 - config_name: en-ur
   features:
   - name: translation
@@ -2149,16 +1908,16 @@ dataset_info:
         - ur
   splits:
   - name: test
-    num_bytes: 538556
+    num_bytes: 538564
     num_examples: 2000
   - name: train
-    num_bytes: 268960696
+    num_bytes: 268961304
     num_examples: 753913
   - name: validation
-    num_bytes: 529308
+    num_bytes: 529316
     num_examples: 2000
-  download_size: 148336044
-  dataset_size: 270028560
+  download_size: 81092186
+  dataset_size: 270029184
 - config_name: en-uz
   features:
   - name: translation
@@ -2169,16 +1928,16 @@ dataset_info:
         - uz
   splits:
   - name: test
-    num_bytes: 408675
+    num_bytes: 408683
     num_examples: 2000
   - name: train
-    num_bytes: 38375290
+    num_bytes: 38375434
     num_examples: 173157
   - name: validation
-    num_bytes: 398853
+    num_bytes: 398861
     num_examples: 2000
-  download_size: 21873536
-  dataset_size: 39182818
+  download_size: 11791643
+  dataset_size: 39182978
 - config_name: en-vi
   features:
   - name: translation
@@ -2189,16 +1948,16 @@ dataset_info:
         - vi
   splits:
   - name: test
-    num_bytes: 192744
+    num_bytes: 192752
     num_examples: 2000
   - name: train
-    num_bytes: 82614470
+    num_bytes: 82615270
     num_examples: 1000000
   - name: validation
-    num_bytes: 194721
+    num_bytes: 194729
     num_examples: 2000
-  download_size: 59250852
-  dataset_size: 83001935
+  download_size: 30647296
+  dataset_size: 83002751
 - config_name: en-wa
   features:
   - name: translation
@@ -2209,16 +1968,16 @@ dataset_info:
         - wa
   splits:
   - name: test
-    num_bytes: 87091
+    num_bytes: 87099
     num_examples: 2000
   - name: train
-    num_bytes: 6085860
+    num_bytes: 6085948
     num_examples: 104496
   - name: validation
-    num_bytes: 87718
+    num_bytes: 87726
     num_examples: 2000
-  download_size: 4512204
-  dataset_size: 6260669
+  download_size: 2119821
+  dataset_size: 6260773
 - config_name: en-xh
   features:
   - name: translation
@@ -2229,16 +1988,16 @@ dataset_info:
         - xh
   splits:
   - name: test
-    num_bytes: 318652
+    num_bytes: 318660
     num_examples: 2000
   - name: train
-    num_bytes: 50606896
+    num_bytes: 50607248
     num_examples: 439671
   - name: validation
-    num_bytes: 315831
+    num_bytes: 315839
     num_examples: 2000
-  download_size: 37519365
-  dataset_size: 51241379
+  download_size: 20503199
+  dataset_size: 51241747
 - config_name: en-yi
   features:
   - name: translation
@@ -2249,16 +2008,16 @@ dataset_info:
         - yi
   splits:
   - name: test
-    num_bytes: 96482
+    num_bytes: 96490
     num_examples: 2000
   - name: train
-    num_bytes: 1275127
+    num_bytes: 1275143
     num_examples: 15010
   - name: validation
-    num_bytes: 99818
+    num_bytes: 99826
     num_examples: 2000
-  download_size: 650530
-  dataset_size: 1471427
+  download_size: 284031
+  dataset_size: 1471459
 - config_name: en-yo
   features:
   - name: translation
@@ -2269,10 +2028,10 @@ dataset_info:
         - yo
   splits:
   - name: train
-    num_bytes: 979753
+    num_bytes: 979769
     num_examples: 10375
-  download_size: 391299
-  dataset_size: 979753
+  download_size: 177540
+  dataset_size: 979769
 - config_name: en-zh
   features:
   - name: translation
@@ -2283,16 +2042,16 @@ dataset_info:
         - zh
   splits:
   - name: test
-    num_bytes: 511364
+    num_bytes: 511372
     num_examples: 2000
   - name: train
-    num_bytes: 200062183
+    num_bytes: 200062983
     num_examples: 1000000
   - name: validation
-    num_bytes: 512356
+    num_bytes: 512364
     num_examples: 2000
-  download_size: 143414756
-  dataset_size: 201085903
+  download_size: 83265500
+  dataset_size: 201086719
 - config_name: en-zu
   features:
   - name: translation
@@ -2303,16 +2062,142 @@ dataset_info:
         - zu
   splits:
   - name: test
-    num_bytes: 117510
+    num_bytes: 117518
     num_examples: 2000
   - name: train
-    num_bytes: 2799558
+    num_bytes: 2799590
     num_examples: 38616
   - name: validation
-    num_bytes: 120133
+    num_bytes: 120141
+    num_examples: 2000
+  download_size: 889951
+  dataset_size: 3037249
+- config_name: ar-de
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - ar
+        - de
+  splits:
+  - name: test
+    num_bytes: 238599
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 238599
+- config_name: ar-fr
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - ar
+        - fr
+  splits:
+  - name: test
+    num_bytes: 547382
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 547382
+- config_name: ar-nl
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - ar
+        - nl
+  splits:
+  - name: test
+    num_bytes: 212936
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 212936
+- config_name: ar-ru
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - ar
+        - ru
+  splits:
+  - name: test
+    num_bytes: 808270
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 808270
+- config_name: ar-zh
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - ar
+        - zh
+  splits:
+  - name: test
+    num_bytes: 713412
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 713412
+- config_name: de-fr
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - de
+        - fr
+  splits:
+  - name: test
+    num_bytes: 458746
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 458746
+- config_name: de-nl
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - de
+        - nl
+  splits:
+  - name: test
+    num_bytes: 403886
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 403886
+- config_name: de-ru
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - de
+        - ru
+  splits:
+  - name: test
+    num_bytes: 315779
     num_examples: 2000
-  download_size: 1918443
-  dataset_size: 3037201
+  download_size: 2556791
+  dataset_size: 315779
+- config_name: de-zh
+  features:
+  - name: translation
+    dtype:
+      translation:
+        languages:
+        - de
+        - zh
+  splits:
+  - name: test
+    num_bytes: 280397
+    num_examples: 2000
+  download_size: 2556791
+  dataset_size: 280397
 - config_name: fr-nl
   features:
   - name: translation
@@ -2323,10 +2208,10 @@ dataset_info:
         - nl
   splits:
   - name: test
-    num_bytes: 368638
+    num_bytes: 368646
     num_examples: 2000
-  download_size: 261290
-  dataset_size: 368638
+  download_size: 2556791
+  dataset_size: 368646
 - config_name: fr-ru
   features:
   - name: translation
@@ -2337,10 +2222,10 @@ dataset_info:
         - ru
   splits:
   - name: test
-    num_bytes: 732716
+    num_bytes: 732724
     num_examples: 2000
-  download_size: 426179
-  dataset_size: 732716
+  download_size: 2556791
+  dataset_size: 732724
 - config_name: fr-zh
   features:
   - name: translation
@@ -2351,10 +2236,10 @@ dataset_info:
         - zh
   splits:
   - name: test
-    num_bytes: 619386
+    num_bytes: 619394
     num_examples: 2000
-  download_size: 418661
-  dataset_size: 619386
+  download_size: 2556791
+  dataset_size: 619394
 - config_name: nl-ru
   features:
   - name: translation
@@ -2365,10 +2250,10 @@ dataset_info:
         - ru
   splits:
   - name: test
-    num_bytes: 256059
+    num_bytes: 256067
     num_examples: 2000
-  download_size: 168666
-  dataset_size: 256059
+  download_size: 2556791
+  dataset_size: 256067
 - config_name: nl-zh
   features:
   - name: translation
@@ -2379,10 +2264,10 @@ dataset_info:
         - zh
   splits:
   - name: test
-    num_bytes: 183633
+    num_bytes: 183641
     num_examples: 2000
-  download_size: 146191
-  dataset_size: 183633
+  download_size: 2556791
+  dataset_size: 183641
 - config_name: ru-zh
   features:
   - name: translation
@@ -2393,846 +2278,128 @@ dataset_info:
         - zh
   splits:
   - name: test
-    num_bytes: 916106
+    num_bytes: 916114
     num_examples: 2000
-  download_size: 534430
-  dataset_size: 916106
-configs:
-- config_name: af-en
-  data_files:
-  - split: test
-    path: af-en/test-*
-  - split: train
-    path: af-en/train-*
-  - split: validation
-    path: af-en/validation-*
-- config_name: am-en
-  data_files:
-  - split: test
-    path: am-en/test-*
-  - split: train
-    path: am-en/train-*
-  - split: validation
-    path: am-en/validation-*
-- config_name: an-en
-  data_files:
-  - split: train
-    path: an-en/train-*
-- config_name: ar-de
-  data_files:
-  - split: test
-    path: ar-de/test-*
-- config_name: ar-en
-  data_files:
-  - split: test
-    path: ar-en/test-*
-  - split: train
-    path: ar-en/train-*
-  - split: validation
-    path: ar-en/validation-*
-- config_name: ar-fr
-  data_files:
-  - split: test
-    path: ar-fr/test-*
-- config_name: ar-nl
-  data_files:
-  - split: test
-    path: ar-nl/test-*
-- config_name: ar-ru
-  data_files:
-  - split: test
-    path: ar-ru/test-*
-- config_name: ar-zh
-  data_files:
-  - split: test
-    path: ar-zh/test-*
-- config_name: as-en
-  data_files:
-  - split: test
-    path: as-en/test-*
-  - split: train
-    path: as-en/train-*
-  - split: validation
-    path: as-en/validation-*
-- config_name: az-en
-  data_files:
-  - split: test
-    path: az-en/test-*
-  - split: train
-    path: az-en/train-*
-  - split: validation
-    path: az-en/validation-*
-- config_name: be-en
-  data_files:
-  - split: test
-    path: be-en/test-*
-  - split: train
-    path: be-en/train-*
-  - split: validation
-    path: be-en/validation-*
-- config_name: bg-en
-  data_files:
-  - split: test
-    path: bg-en/test-*
-  - split: train
-    path: bg-en/train-*
-  - split: validation
-    path: bg-en/validation-*
-- config_name: bn-en
-  data_files:
-  - split: test
-    path: bn-en/test-*
-  - split: train
-    path: bn-en/train-*
-  - split: validation
-    path: bn-en/validation-*
-- config_name: br-en
-  data_files:
-  - split: test
-    path: br-en/test-*
-  - split: train
-    path: br-en/train-*
-  - split: validation
-    path: br-en/validation-*
-- config_name: bs-en
-  data_files:
-  - split: test
-    path: bs-en/test-*
-  - split: train
-    path: bs-en/train-*
-  - split: validation
-    path: bs-en/validation-*
-- config_name: ca-en
-  data_files:
-  - split: test
-    path: ca-en/test-*
-  - split: train
-    path: ca-en/train-*
-  - split: validation
-    path: ca-en/validation-*
-- config_name: cs-en
-  data_files:
-  - split: test
-    path: cs-en/test-*
-  - split: train
-    path: cs-en/train-*
-  - split: validation
-    path: cs-en/validation-*
-- config_name: cy-en
-  data_files:
-  - split: test
-    path: cy-en/test-*
-  - split: train
-    path: cy-en/train-*
-  - split: validation
-    path: cy-en/validation-*
-- config_name: da-en
-  data_files:
-  - split: test
-    path: da-en/test-*
-  - split: train
-    path: da-en/train-*
-  - split: validation
-    path: da-en/validation-*
-- config_name: de-en
-  data_files:
-  - split: test
-    path: de-en/test-*
-  - split: train
-    path: de-en/train-*
-  - split: validation
-    path: de-en/validation-*
-- config_name: de-fr
-  data_files:
-  - split: test
-    path: de-fr/test-*
-- config_name: de-nl
-  data_files:
-  - split: test
-    path: de-nl/test-*
-- config_name: de-ru
-  data_files:
-  - split: test
-    path: de-ru/test-*
-- config_name: de-zh
-  data_files:
-  - split: test
-    path: de-zh/test-*
-- config_name: dz-en
-  data_files:
-  - split: train
-    path: dz-en/train-*
-- config_name: el-en
-  data_files:
-  - split: test
-    path: el-en/test-*
-  - split: train
-    path: el-en/train-*
-  - split: validation
-    path: el-en/validation-*
-- config_name: en-eo
-  data_files:
-  - split: test
-    path: en-eo/test-*
-  - split: train
-    path: en-eo/train-*
-  - split: validation
-    path: en-eo/validation-*
-- config_name: en-es
-  data_files:
-  - split: test
-    path: en-es/test-*
-  - split: train
-    path: en-es/train-*
-  - split: validation
-    path: en-es/validation-*
-- config_name: en-et
-  data_files:
-  - split: test
-    path: en-et/test-*
-  - split: train
-    path: en-et/train-*
-  - split: validation
-    path: en-et/validation-*
-- config_name: en-eu
-  data_files:
-  - split: test
-    path: en-eu/test-*
-  - split: train
-    path: en-eu/train-*
-  - split: validation
-    path: en-eu/validation-*
-- config_name: en-fa
-  data_files:
-  - split: test
-    path: en-fa/test-*
-  - split: train
-    path: en-fa/train-*
-  - split: validation
-    path: en-fa/validation-*
-- config_name: en-fi
-  data_files:
-  - split: test
-    path: en-fi/test-*
-  - split: train
-    path: en-fi/train-*
-  - split: validation
-    path: en-fi/validation-*
-- config_name: en-fr
-  data_files:
-  - split: test
-    path: en-fr/test-*
-  - split: train
-    path: en-fr/train-*
-  - split: validation
-    path: en-fr/validation-*
-- config_name: en-fy
-  data_files:
-  - split: test
-    path: en-fy/test-*
-  - split: train
-    path: en-fy/train-*
-  - split: validation
-    path: en-fy/validation-*
-- config_name: en-ga
-  data_files:
-  - split: test
-    path: en-ga/test-*
-  - split: train
-    path: en-ga/train-*
-  - split: validation
-    path: en-ga/validation-*
-- config_name: en-gd
-  data_files:
-  - split: test
-    path: en-gd/test-*
-  - split: train
-    path: en-gd/train-*
-  - split: validation
-    path: en-gd/validation-*
-- config_name: en-gl
-  data_files:
-  - split: test
-    path: en-gl/test-*
-  - split: train
-    path: en-gl/train-*
-  - split: validation
-    path: en-gl/validation-*
-- config_name: en-gu
-  data_files:
-  - split: test
-    path: en-gu/test-*
-  - split: train
-    path: en-gu/train-*
-  - split: validation
-    path: en-gu/validation-*
-- config_name: en-ha
-  data_files:
-  - split: test
-    path: en-ha/test-*
-  - split: train
-    path: en-ha/train-*
-  - split: validation
-    path: en-ha/validation-*
-- config_name: en-he
-  data_files:
-  - split: test
-    path: en-he/test-*
-  - split: train
-    path: en-he/train-*
-  - split: validation
-    path: en-he/validation-*
-- config_name: en-hi
-  data_files:
-  - split: test
-    path: en-hi/test-*
-  - split: train
-    path: en-hi/train-*
-  - split: validation
-    path: en-hi/validation-*
-- config_name: en-hr
-  data_files:
-  - split: test
-    path: en-hr/test-*
-  - split: train
-    path: en-hr/train-*
-  - split: validation
-    path: en-hr/validation-*
-- config_name: en-hu
-  data_files:
-  - split: test
-    path: en-hu/test-*
-  - split: train
-    path: en-hu/train-*
-  - split: validation
-    path: en-hu/validation-*
-- config_name: en-hy
-  data_files:
-  - split: train
-    path: en-hy/train-*
-- config_name: en-id
-  data_files:
-  - split: test
-    path: en-id/test-*
-  - split: train
-    path: en-id/train-*
-  - split: validation
-    path: en-id/validation-*
-- config_name: en-ig
-  data_files:
-  - split: test
-    path: en-ig/test-*
-  - split: train
-    path: en-ig/train-*
-  - split: validation
-    path: en-ig/validation-*
-- config_name: en-is
-  data_files:
-  - split: test
-    path: en-is/test-*
-  - split: train
-    path: en-is/train-*
-  - split: validation
-    path: en-is/validation-*
-- config_name: en-it
-  data_files:
-  - split: test
-    path: en-it/test-*
-  - split: train
-    path: en-it/train-*
-  - split: validation
-    path: en-it/validation-*
-- config_name: en-ja
-  data_files:
-  - split: test
-    path: en-ja/test-*
-  - split: train
-    path: en-ja/train-*
-  - split: validation
-    path: en-ja/validation-*
-- config_name: en-ka
-  data_files:
-  - split: test
-    path: en-ka/test-*
-  - split: train
-    path: en-ka/train-*
-  - split: validation
-    path: en-ka/validation-*
-- config_name: en-kk
-  data_files:
-  - split: test
-    path: en-kk/test-*
-  - split: train
-    path: en-kk/train-*
-  - split: validation
-    path: en-kk/validation-*
-- config_name: en-km
-  data_files:
-  - split: test
-    path: en-km/test-*
-  - split: train
-    path: en-km/train-*
-  - split: validation
-    path: en-km/validation-*
-- config_name: en-kn
-  data_files:
-  - split: test
-    path: en-kn/test-*
-  - split: train
-    path: en-kn/train-*
-  - split: validation
-    path: en-kn/validation-*
-- config_name: en-ko
-  data_files:
-  - split: test
-    path: en-ko/test-*
-  - split: train
-    path: en-ko/train-*
-  - split: validation
-    path: en-ko/validation-*
-- config_name: en-ku
-  data_files:
-  - split: test
-    path: en-ku/test-*
-  - split: train
-    path: en-ku/train-*
-  - split: validation
-    path: en-ku/validation-*
-- config_name: en-ky
-  data_files:
-  - split: test
-    path: en-ky/test-*
-  - split: train
-    path: en-ky/train-*
-  - split: validation
-    path: en-ky/validation-*
-- config_name: en-li
-  data_files:
-  - split: test
-    path: en-li/test-*
-  - split: train
-    path: en-li/train-*
-  - split: validation
-    path: en-li/validation-*
-- config_name: en-lt
-  data_files:
-  - split: test
-    path: en-lt/test-*
-  - split: train
-    path: en-lt/train-*
-  - split: validation
-    path: en-lt/validation-*
-- config_name: en-lv
-  data_files:
-  - split: test
-    path: en-lv/test-*
-  - split: train
-    path: en-lv/train-*
-  - split: validation
-    path: en-lv/validation-*
-- config_name: en-mg
-  data_files:
-  - split: test
-    path: en-mg/test-*
-  - split: train
-    path: en-mg/train-*
-  - split: validation
-    path: en-mg/validation-*
-- config_name: en-mk
-  data_files:
-  - split: test
-    path: en-mk/test-*
-  - split: train
-    path: en-mk/train-*
-  - split: validation
-    path: en-mk/validation-*
-- config_name: en-ml
-  data_files:
-  - split: test
-    path: en-ml/test-*
-  - split: train
-    path: en-ml/train-*
-  - split: validation
-    path: en-ml/validation-*
-- config_name: en-mn
-  data_files:
-  - split: train
-    path: en-mn/train-*
-- config_name: en-mr
-  data_files:
-  - split: test
-    path: en-mr/test-*
-  - split: train
-    path: en-mr/train-*
-  - split: validation
-    path: en-mr/validation-*
-- config_name: en-ms
-  data_files:
-  - split: test
-    path: en-ms/test-*
-  - split: train
-    path: en-ms/train-*
-  - split: validation
-    path: en-ms/validation-*
-- config_name: en-mt
-  data_files:
-  - split: test
-    path: en-mt/test-*
-  - split: train
-    path: en-mt/train-*
-  - split: validation
-    path: en-mt/validation-*
-- config_name: en-my
-  data_files:
-  - split: test
-    path: en-my/test-*
-  - split: train
-    path: en-my/train-*
-  - split: validation
-    path: en-my/validation-*
-- config_name: en-nb
-  data_files:
-  - split: test
-    path: en-nb/test-*
-  - split: train
-    path: en-nb/train-*
-  - split: validation
-    path: en-nb/validation-*
-- config_name: en-ne
-  data_files:
-  - split: test
-    path: en-ne/test-*
-  - split: train
-    path: en-ne/train-*
-  - split: validation
-    path: en-ne/validation-*
-- config_name: en-nl
-  data_files:
-  - split: test
-    path: en-nl/test-*
-  - split: train
-    path: en-nl/train-*
-  - split: validation
-    path: en-nl/validation-*
-- config_name: en-nn
-  data_files:
-  - split: test
-    path: en-nn/test-*
-  - split: train
-    path: en-nn/train-*
-  - split: validation
-    path: en-nn/validation-*
-- config_name: en-no
-  data_files:
-  - split: test
-    path: en-no/test-*
-  - split: train
-    path: en-no/train-*
-  - split: validation
-    path: en-no/validation-*
-- config_name: en-oc
-  data_files:
-  - split: test
-    path: en-oc/test-*
-  - split: train
-    path: en-oc/train-*
-  - split: validation
-    path: en-oc/validation-*
-- config_name: en-or
-  data_files:
-  - split: test
-    path: en-or/test-*
-  - split: train
-    path: en-or/train-*
-  - split: validation
-    path: en-or/validation-*
-- config_name: en-pa
-  data_files:
-  - split: test
-    path: en-pa/test-*
-  - split: train
-    path: en-pa/train-*
-  - split: validation
-    path: en-pa/validation-*
-- config_name: en-pl
-  data_files:
-  - split: test
-    path: en-pl/test-*
-  - split: train
-    path: en-pl/train-*
-  - split: validation
-    path: en-pl/validation-*
-- config_name: en-ps
-  data_files:
-  - split: test
-    path: en-ps/test-*
-  - split: train
-    path: en-ps/train-*
-  - split: validation
-    path: en-ps/validation-*
-- config_name: en-pt
-  data_files:
-  - split: test
-    path: en-pt/test-*
-  - split: train
-    path: en-pt/train-*
-  - split: validation
-    path: en-pt/validation-*
-- config_name: en-ro
-  data_files:
-  - split: test
-    path: en-ro/test-*
-  - split: train
-    path: en-ro/train-*
-  - split: validation
-    path: en-ro/validation-*
-- config_name: en-ru
-  data_files:
-  - split: test
-    path: en-ru/test-*
-  - split: train
-    path: en-ru/train-*
-  - split: validation
-    path: en-ru/validation-*
-- config_name: en-rw
-  data_files:
-  - split: test
-    path: en-rw/test-*
-  - split: train
-    path: en-rw/train-*
-  - split: validation
-    path: en-rw/validation-*
-- config_name: en-se
-  data_files:
-  - split: test
-    path: en-se/test-*
-  - split: train
-    path: en-se/train-*
-  - split: validation
-    path: en-se/validation-*
-- config_name: en-sh
-  data_files:
-  - split: test
-    path: en-sh/test-*
-  - split: train
-    path: en-sh/train-*
-  - split: validation
-    path: en-sh/validation-*
-- config_name: en-si
-  data_files:
-  - split: test
-    path: en-si/test-*
-  - split: train
-    path: en-si/train-*
-  - split: validation
-    path: en-si/validation-*
-- config_name: en-sk
-  data_files:
-  - split: test
-    path: en-sk/test-*
-  - split: train
-    path: en-sk/train-*
-  - split: validation
-    path: en-sk/validation-*
-- config_name: en-sl
-  data_files:
-  - split: test
-    path: en-sl/test-*
-  - split: train
-    path: en-sl/train-*
-  - split: validation
-    path: en-sl/validation-*
-- config_name: en-sq
-  data_files:
-  - split: test
-    path: en-sq/test-*
-  - split: train
-    path: en-sq/train-*
-  - split: validation
-    path: en-sq/validation-*
-- config_name: en-sr
-  data_files:
-  - split: test
-    path: en-sr/test-*
-  - split: train
-    path: en-sr/train-*
-  - split: validation
-    path: en-sr/validation-*
-- config_name: en-sv
-  data_files:
-  - split: test
-    path: en-sv/test-*
-  - split: train
-    path: en-sv/train-*
-  - split: validation
-    path: en-sv/validation-*
-- config_name: en-ta
-  data_files:
-  - split: test
-    path: en-ta/test-*
-  - split: train
-    path: en-ta/train-*
-  - split: validation
-    path: en-ta/validation-*
-- config_name: en-te
-  data_files:
-  - split: test
-    path: en-te/test-*
-  - split: train
-    path: en-te/train-*
-  - split: validation
-    path: en-te/validation-*
-- config_name: en-tg
-  data_files:
-  - split: test
-    path: en-tg/test-*
-  - split: train
-    path: en-tg/train-*
-  - split: validation
-    path: en-tg/validation-*
-- config_name: en-th
-  data_files:
-  - split: test
-    path: en-th/test-*
-  - split: train
-    path: en-th/train-*
-  - split: validation
-    path: en-th/validation-*
-- config_name: en-tk
-  data_files:
-  - split: test
-    path: en-tk/test-*
-  - split: train
-    path: en-tk/train-*
-  - split: validation
-    path: en-tk/validation-*
-- config_name: en-tr
-  data_files:
-  - split: test
-    path: en-tr/test-*
-  - split: train
-    path: en-tr/train-*
-  - split: validation
-    path: en-tr/validation-*
-- config_name: en-tt
-  data_files:
-  - split: test
-    path: en-tt/test-*
-  - split: train
-    path: en-tt/train-*
-  - split: validation
-    path: en-tt/validation-*
-- config_name: en-ug
-  data_files:
-  - split: test
-    path: en-ug/test-*
-  - split: train
-    path: en-ug/train-*
-  - split: validation
-    path: en-ug/validation-*
-- config_name: en-uk
-  data_files:
-  - split: test
-    path: en-uk/test-*
-  - split: train
-    path: en-uk/train-*
-  - split: validation
-    path: en-uk/validation-*
-- config_name: en-ur
-  data_files:
-  - split: test
-    path: en-ur/test-*
-  - split: train
-    path: en-ur/train-*
-  - split: validation
-    path: en-ur/validation-*
-- config_name: en-uz
-  data_files:
-  - split: test
-    path: en-uz/test-*
-  - split: train
-    path: en-uz/train-*
-  - split: validation
-    path: en-uz/validation-*
-- config_name: en-vi
-  data_files:
-  - split: test
-    path: en-vi/test-*
-  - split: train
-    path: en-vi/train-*
-  - split: validation
-    path: en-vi/validation-*
-- config_name: en-wa
-  data_files:
-  - split: test
-    path: en-wa/test-*
-  - split: train
-    path: en-wa/train-*
-  - split: validation
-    path: en-wa/validation-*
-- config_name: en-xh
-  data_files:
-  - split: test
-    path: en-xh/test-*
-  - split: train
-    path: en-xh/train-*
-  - split: validation
-    path: en-xh/validation-*
-- config_name: en-yi
-  data_files:
-  - split: test
-    path: en-yi/test-*
-  - split: train
-    path: en-yi/train-*
-  - split: validation
-    path: en-yi/validation-*
-- config_name: en-yo
-  data_files:
-  - split: train
-    path: en-yo/train-*
-- config_name: en-zh
-  data_files:
-  - split: test
-    path: en-zh/test-*
-  - split: train
-    path: en-zh/train-*
-  - split: validation
-    path: en-zh/validation-*
-- config_name: en-zu
-  data_files:
-  - split: test
-    path: en-zu/test-*
-  - split: train
-    path: en-zu/train-*
-  - split: validation
-    path: en-zu/validation-*
-- config_name: fr-nl
-  data_files:
-  - split: test
-    path: fr-nl/test-*
-- config_name: fr-ru
-  data_files:
-  - split: test
-    path: fr-ru/test-*
-- config_name: fr-zh
-  data_files:
-  - split: test
-    path: fr-zh/test-*
-- config_name: nl-ru
-  data_files:
-  - split: test
-    path: nl-ru/test-*
-- config_name: nl-zh
-  data_files:
-  - split: test
-    path: nl-zh/test-*
-- config_name: ru-zh
-  data_files:
-  - split: test
-    path: ru-zh/test-*
+  download_size: 2556791
+  dataset_size: 916114
+config_names:
+- af-en
+- am-en
+- an-en
+- ar-de
+- ar-en
+- ar-fr
+- ar-nl
+- ar-ru
+- ar-zh
+- as-en
+- az-en
+- be-en
+- bg-en
+- bn-en
+- br-en
+- bs-en
+- ca-en
+- cs-en
+- cy-en
+- da-en
+- de-en
+- de-fr
+- de-nl
+- de-ru
+- de-zh
+- dz-en
+- el-en
+- en-eo
+- en-es
+- en-et
+- en-eu
+- en-fa
+- en-fi
+- en-fr
+- en-fy
+- en-ga
+- en-gd
+- en-gl
+- en-gu
+- en-ha
+- en-he
+- en-hi
+- en-hr
+- en-hu
+- en-hy
+- en-id
+- en-ig
+- en-is
+- en-it
+- en-ja
+- en-ka
+- en-kk
+- en-km
+- en-kn
+- en-ko
+- en-ku
+- en-ky
+- en-li
+- en-lt
+- en-lv
+- en-mg
+- en-mk
+- en-ml
+- en-mn
+- en-mr
+- en-ms
+- en-mt
+- en-my
+- en-nb
+- en-ne
+- en-nl
+- en-nn
+- en-no
+- en-oc
+- en-or
+- en-pa
+- en-pl
+- en-ps
+- en-pt
+- en-ro
+- en-ru
+- en-rw
+- en-se
+- en-sh
+- en-si
+- en-sk
+- en-sl
+- en-sq
+- en-sr
+- en-sv
+- en-ta
+- en-te
+- en-tg
+- en-th
+- en-tk
+- en-tr
+- en-tt
+- en-ug
+- en-uk
+- en-ur
+- en-uz
+- en-vi
+- en-wa
+- en-xh
+- en-yi
+- en-yo
+- en-zh
+- en-zu
+- fr-nl
+- fr-ru
+- fr-zh
+- nl-ru
+- nl-zh
+- ru-zh
 ---
 
-# Dataset Card for OPUS-100
+# Dataset Card for Opus100
 
 ## Table of Contents
 - [Dataset Description](#dataset-description)
@@ -3260,23 +2427,19 @@ configs:
 
 ## Dataset Description
 
-- **Homepage:** https://opus.nlpl.eu/OPUS-100
-- **Repository:** https://github.com/EdinburghNLP/opus-100-corpus
-- **Paper:** https://arxiv.org/abs/2004.11867
-- **Paper:** https://aclanthology.org/L10-1473/
-- **Leaderboard:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
-- **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+- **Homepage:** [Link](http://opus.nlpl.eu/opus-100.php) 
+- **Repository:** [GitHub](https://github.com/EdinburghNLP/opus-100-corpus)
+- **Paper:** [ARXIV](https://arxiv.org/abs/2004.11867)
+- **Leaderboard:** 
+- **Point of Contact:** 
 
 ### Dataset Summary
 
-OPUS-100 is an English-centric multilingual corpus covering 100 languages.
-
-OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side. The corpus covers 100 languages (including English).
-The languages were selected based on the volume of parallel data available in OPUS.
+OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side. The corpus covers 100 languages (including English). Selected the languages based on the volume of parallel data available in OPUS.
 
 ### Supported Tasks and Leaderboards
 
-Translation.
+[More Information Needed]
 
 ### Languages
 
@@ -3288,16 +2451,15 @@ OPUS-100 contains approximately 55M sentence pairs. Of the 99 language pairs, 44
 
 ```
 {
-  "translation": {
-    "ca": "El departament de bombers té el seu propi equip d'investigació.",
-    "en": "Well, the fire department has its own investigative unit."
-  }
+  "ca": "El departament de bombers té el seu propi equip d'investigació.",
+  "en": "Well, the fire department has its own investigative unit."
 }
 ```
 
 ### Data Fields
 
-- `translation` (`dict`): Parallel sentences for the pair of languages.
+- `src_tag`: `string` text in source language
+- `tgt_tag`: `string` translation of source language in target language
 
 ### Data Splits
 
@@ -3361,49 +2523,14 @@ The dataset is split into training, development, and test portions. Data was pre
 
 ### Citation Information
 
-If you use this corpus, please cite the paper:
-```bibtex
-@inproceedings{zhang-etal-2020-improving,
-    title = "Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation",
-    author = "Zhang, Biao  and
-      Williams, Philip  and
-      Titov, Ivan  and
-      Sennrich, Rico",
-    editor = "Jurafsky, Dan  and
-      Chai, Joyce  and
-      Schluter, Natalie  and
-      Tetreault, Joel",
-    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
-    month = jul,
-    year = "2020",
-    address = "Online",
-    publisher = "Association for Computational Linguistics",
-    url = "https://aclanthology.org/2020.acl-main.148",
-    doi = "10.18653/v1/2020.acl-main.148",
-    pages = "1628--1639",
-}
 ```
-and, please, also acknowledge OPUS:
-```bibtex
-@inproceedings{tiedemann-2012-parallel,
-    title = "Parallel Data, Tools and Interfaces in {OPUS}",
-    author = {Tiedemann, J{\"o}rg},
-    editor = "Calzolari, Nicoletta  and
-      Choukri, Khalid  and
-      Declerck, Thierry  and
-      Do{\u{g}}an, Mehmet U{\u{g}}ur  and
-      Maegaard, Bente  and
-      Mariani, Joseph  and
-      Moreno, Asuncion  and
-      Odijk, Jan  and
-      Piperidis, Stelios",
-    booktitle = "Proceedings of the Eighth International Conference on Language Resources and Evaluation ({LREC}'12)",
-    month = may,
-    year = "2012",
-    address = "Istanbul, Turkey",
-    publisher = "European Language Resources Association (ELRA)",
-    url = "http://www.lrec-conf.org/proceedings/lrec2012/pdf/463_Paper.pdf",
-    pages = "2214--2218",
+@misc{zhang2020improving,
+      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation}, 
+      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},
+      year={2020},
+      eprint={2004.11867},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
 }
 ```
 
diff --git a/af-en/test-00000-of-00001.parquet b/af-en/test-00000-of-00001.parquet
deleted file mode 100644
index 35f05b577dfc4bdd06ffeceec5ae89945e164f82..0000000000000000000000000000000000000000
--- a/af-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d27bde98a04294cbfb2743b35040c06ee058acc1cef856343d0f961e616a4bef
-size 107782
diff --git a/af-en/train-00000-of-00001.parquet b/af-en/train-00000-of-00001.parquet
deleted file mode 100644
index fa92662f62be0e59f1303427abebdeec682bc356..0000000000000000000000000000000000000000
--- a/af-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:16f1a4225efd61fe1ce606b36e1feb9970a4b5f64093ba1814b73446e221e6e2
-size 14640759
diff --git a/af-en/validation-00000-of-00001.parquet b/af-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 0ee851cda9e961b7a88d3a26939fbde987678fc5..0000000000000000000000000000000000000000
--- a/af-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fef36df6d9238fdd543838a184d84dc3d4ba5bc011d352e6c5f4d4bad5705e1d
-size 104256
diff --git a/am-en/test-00000-of-00001.parquet b/am-en/test-00000-of-00001.parquet
deleted file mode 100644
index 07b2840613473a0bb198c1d611fed1d65f1bfa72..0000000000000000000000000000000000000000
--- a/am-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1a330ef37d29e786be58221a25f1f362cff7ced3a4f0d34ce738ca1b69596c36
-size 317531
diff --git a/am-en/train-00000-of-00001.parquet b/am-en/train-00000-of-00001.parquet
deleted file mode 100644
index 03ee1b2a5635d1b65af713e0faf3aa471ccf63fc..0000000000000000000000000000000000000000
--- a/am-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b2fa6bdf344846c59dca4cbdcaa649811b2dbedb39966e2a779d7e2c8d640e09
-size 12006927
diff --git a/am-en/validation-00000-of-00001.parquet b/am-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 2b28c4fa9951dc109408d948840da9bdf7c9eeb3..0000000000000000000000000000000000000000
--- a/am-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d682db4109033ff3cc0bc6575c3c01b7af10d311a3350e4f6f116d76f5802ef8
-size 305573
diff --git a/an-en/train-00000-of-00001.parquet b/an-en/train-00000-of-00001.parquet
deleted file mode 100644
index 8ec64f046740d694e71b5b5d71308fbfc860184c..0000000000000000000000000000000000000000
--- a/an-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fd2d50748d4d37996523dc5751acf2242b235c12cc1c11e7ccf831acb4ac79a2
-size 232976
diff --git a/ar-de/test-00000-of-00001.parquet b/ar-de/test-00000-of-00001.parquet
deleted file mode 100644
index 43cf5b2e5e4d4c86bb8a4af7e8d0bada5d13d90c..0000000000000000000000000000000000000000
--- a/ar-de/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c330c428a4668a6b50033967dd6e6b9a98338e5d0265e2b37ac30cd1c9f73a21
-size 161557
diff --git a/ar-en/test-00000-of-00001.parquet b/ar-en/test-00000-of-00001.parquet
deleted file mode 100644
index 434c088bbb2e5c3946cae15b0575d95c92124007..0000000000000000000000000000000000000000
--- a/ar-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:63d58413e77fa96140a7bfaafdf444fdb2caca2c33013ebd8242aeaf536e785d
-size 214340
diff --git a/ar-en/train-00000-of-00001.parquet b/ar-en/train-00000-of-00001.parquet
deleted file mode 100644
index 22cb89c0b35cd96e1e956b875a377eaede61279a..0000000000000000000000000000000000000000
--- a/ar-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c8e6b1a3fbe57fbdd126dedc7d16d1142fd8a9534ae4d76141d478ab78431369
-size 99293243
diff --git a/ar-en/validation-00000-of-00001.parquet b/ar-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 1e45b22ada911e8d93abe5b2c823d7709fe163cb..0000000000000000000000000000000000000000
--- a/ar-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4a226fa88468d5e40fe2482bc85a8745d07629d1f6af9e48e50cb237a889b45d
-size 979231
diff --git a/ar-fr/test-00000-of-00001.parquet b/ar-fr/test-00000-of-00001.parquet
deleted file mode 100644
index 9425f6e8493885e1970169bd78091f051e2daead..0000000000000000000000000000000000000000
--- a/ar-fr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1a8d248a1a6d641f943dcbbe3ef17117b18f99ad5aaae421f704ced1e4172dbb
-size 334226
diff --git a/ar-nl/test-00000-of-00001.parquet b/ar-nl/test-00000-of-00001.parquet
deleted file mode 100644
index 76e10e85f9187414bb189b1695fdb3c420b2db3a..0000000000000000000000000000000000000000
--- a/ar-nl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d49496c69c6645e1a2fba57b85118624f67e82d8b7cceda9e0998feebf4786a7
-size 144863
diff --git a/ar-ru/test-00000-of-00001.parquet b/ar-ru/test-00000-of-00001.parquet
deleted file mode 100644
index 1c6fc5172559f431fd454a46d3aa93758fb65f12..0000000000000000000000000000000000000000
--- a/ar-ru/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8ed225a0e6832f1cf3f456d5e69082b5fc6fdc64d466bc5456637746cb46361c
-size 441536
diff --git a/ar-zh/test-00000-of-00001.parquet b/ar-zh/test-00000-of-00001.parquet
deleted file mode 100644
index 2fada57e71c42f932a502179b3633bea129a66dc..0000000000000000000000000000000000000000
--- a/ar-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2049f357507db6c5f820a8b5f9bb9a1d635d54f7cd5163caa468aed6701ac3e9
-size 438598
diff --git a/as-en/test-00000-of-00001.parquet b/as-en/test-00000-of-00001.parquet
deleted file mode 100644
index a6f50d65539db1d8ec752d49a95b6ea2aeed4ece..0000000000000000000000000000000000000000
--- a/as-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c491d88d77400dd21ec00c3513380eed57376446d97df0fe76d499f451ecd6d9
-size 139237
diff --git a/as-en/train-00000-of-00001.parquet b/as-en/train-00000-of-00001.parquet
deleted file mode 100644
index 0914d08e738645d1b7d2b625d6c1dc3184d8b349..0000000000000000000000000000000000000000
--- a/as-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d4e2073b21fbc148e74c1698667d28b71c905c7d7c67bc3099e90826a894809b
-size 8522321
diff --git a/as-en/validation-00000-of-00001.parquet b/as-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 21498492cbffabd4731b3df0fc38467b798289b8..0000000000000000000000000000000000000000
--- a/as-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ead20ed6fdf6f0a6e5a0381f0b5beee9363b82ebba945807175d391a1d941a76
-size 133058
diff --git a/az-en/test-00000-of-00001.parquet b/az-en/test-00000-of-00001.parquet
deleted file mode 100644
index 3e4a72aa1ff5a3b3781514d405e60dc32011582e..0000000000000000000000000000000000000000
--- a/az-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8c5891c698ea1943dbba0c123b9cb9ef21149a3401f846e81c33cb1d41ea0f1d
-size 241419
diff --git a/az-en/train-00000-of-00001.parquet b/az-en/train-00000-of-00001.parquet
deleted file mode 100644
index b0669ad225e414977d0333b572acbcba476ea566..0000000000000000000000000000000000000000
--- a/az-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8528c9892f8a3c7e779bf1305cc3620525c333b7df4dc4d06e8bafc9e32333f4
-size 34498539
diff --git a/az-en/validation-00000-of-00001.parquet b/az-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 61363443d5a7a86200b5f0d2503286e292ed7993..0000000000000000000000000000000000000000
--- a/az-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3fed50edd2df2d4e5e6d89deea43812d09a1f7bd24ae34572141e6ebfae0d80a
-size 248901
diff --git a/be-en/test-00000-of-00001.parquet b/be-en/test-00000-of-00001.parquet
deleted file mode 100644
index d4481ef60f3c143d8c1660c89d223d5d5370fdda..0000000000000000000000000000000000000000
--- a/be-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c276fb556bcf80ab63cd4b9833f5a94725afb90b4a25b9301cf04f7e5f17f27c
-size 109873
diff --git a/be-en/train-00000-of-00001.parquet b/be-en/train-00000-of-00001.parquet
deleted file mode 100644
index ae30604798043e12d5852a8f9a8e233134302be9..0000000000000000000000000000000000000000
--- a/be-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:14efe07e4e72c4b1b84981076e04b0aa215025a9a508041fa44af2552c05cc52
-size 3587362
diff --git a/be-en/validation-00000-of-00001.parquet b/be-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 937b3166ffdb18133dd5758888ae74b0b3832316..0000000000000000000000000000000000000000
--- a/be-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:be9674ae23b1869be95596b6e3f79e59543e22fc197812d0dc6ab4b7cd7e4bf4
-size 110434
diff --git a/bg-en/test-00000-of-00001.parquet b/bg-en/test-00000-of-00001.parquet
deleted file mode 100644
index 94502b2d6541539e183b68946bc7809b29eda59f..0000000000000000000000000000000000000000
--- a/bg-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:67bba087abf0733c784b460b4fea514af4a5082521deb56770dd000f37c89b14
-size 159517
diff --git a/bg-en/train-00000-of-00001.parquet b/bg-en/train-00000-of-00001.parquet
deleted file mode 100644
index 926700909442b5a116c8f7f5acf9f343823adda4..0000000000000000000000000000000000000000
--- a/bg-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:858ace676a4e5d2265dc8620e27b31bcdb1ffd4c01458ac688547b3447969945
-size 71261839
diff --git a/bg-en/validation-00000-of-00001.parquet b/bg-en/validation-00000-of-00001.parquet
deleted file mode 100644
index d7c5af2d8c9bc557bcef1f072a826910f404573d..0000000000000000000000000000000000000000
--- a/bg-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5ab819c3e55d72a1bbaddc94ed987c3b763e88814e18230cba90a7c05f8b7d14
-size 153954
diff --git a/bn-en/test-00000-of-00001.parquet b/bn-en/test-00000-of-00001.parquet
deleted file mode 100644
index 75ebf9f83087e2c873bdb877178a852a41911758..0000000000000000000000000000000000000000
--- a/bn-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:eb2b30422561374efc69b885c8407c357967fc18fc088ac0a61433e7147e4ed8
-size 279391
diff --git a/bn-en/train-00000-of-00001.parquet b/bn-en/train-00000-of-00001.parquet
deleted file mode 100644
index 14491f771592310a682bee4d0bf36054b2ad7e4b..0000000000000000000000000000000000000000
--- a/bn-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4ebfeb6a3a426ea2ab4ffa9bd8856c602ce99d5cc239e58b570acee54555c662
-size 133525065
diff --git a/bn-en/validation-00000-of-00001.parquet b/bn-en/validation-00000-of-00001.parquet
deleted file mode 100644
index a2043bd8eca0081d559d5f11330c22b0e4964832..0000000000000000000000000000000000000000
--- a/bn-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ee9db834d030e71536a769c3c0041689c8a875b37a8440eff9d6f8cd1f4e794b
-size 272140
diff --git a/br-en/test-00000-of-00001.parquet b/br-en/test-00000-of-00001.parquet
deleted file mode 100644
index 7b709ee721911bdbc7fae794c49fe460cc3f135e..0000000000000000000000000000000000000000
--- a/br-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e9fbbadda5492cbca70e3b66ec39ed3995b524a42d50c14aa842ffc1796b3c07
-size 98772
diff --git a/br-en/train-00000-of-00001.parquet b/br-en/train-00000-of-00001.parquet
deleted file mode 100644
index c4de42b20c9b6b085954ec7b3597a53d27ebd39d..0000000000000000000000000000000000000000
--- a/br-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:49191597d48d035df9d487032ed601677c2c5d6e36945ae5aebf715ca8ff69d7
-size 6681649
diff --git a/br-en/validation-00000-of-00001.parquet b/br-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 8a96ae1068ec2167de918027711e440babe9c01a..0000000000000000000000000000000000000000
--- a/br-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2d6bf558976224fed0e34d0558bb87979b06c72216e66f4e0cc6442ac8eafc8a
-size 101444
diff --git a/bs-en/test-00000-of-00001.parquet b/bs-en/test-00000-of-00001.parquet
deleted file mode 100644
index a11acdc32bf7e59d0f845511e6feeae4759ff4b6..0000000000000000000000000000000000000000
--- a/bs-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b8196388b93e93b615ef5240927fa403b69c9f56b00d40174e905f53ac6449fa
-size 132538
diff --git a/bs-en/train-00000-of-00001.parquet b/bs-en/train-00000-of-00001.parquet
deleted file mode 100644
index 6ae0f0536438253b8674b87150dbf682c45cd442..0000000000000000000000000000000000000000
--- a/bs-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:547dbf7484b93ec0eb8805c4b90618a9501805c6f4f3a56dee19b6c8c72fb626
-size 59245618
diff --git a/bs-en/validation-00000-of-00001.parquet b/bs-en/validation-00000-of-00001.parquet
deleted file mode 100644
index ce268752b6ff86a7f8cc5c21450feb41e385def9..0000000000000000000000000000000000000000
--- a/bs-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2208352a0c5b298ffc0c0a318700b256e7b823ed7b8a071d634b3f21765acd85
-size 136247
diff --git a/ca-en/test-00000-of-00001.parquet b/ca-en/test-00000-of-00001.parquet
deleted file mode 100644
index e5875800cdbcd209189af247bcb27c9c74c04cfd..0000000000000000000000000000000000000000
--- a/ca-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2de4cc382ea1253d688483a6b12bb02a3bea9ba77489bd7040148d0f0be4c75f
-size 155369
diff --git a/ca-en/train-00000-of-00001.parquet b/ca-en/train-00000-of-00001.parquet
deleted file mode 100644
index 2e63ed1c8f4cd523f50ce49f77905d8a785c2802..0000000000000000000000000000000000000000
--- a/ca-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:35774c65682037478a4b1c72159cfa7606d8b49c38bab3fd5425a6f32324b353
-size 68118730
diff --git a/ca-en/validation-00000-of-00001.parquet b/ca-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 5c90e560b637866f960b1f82c2d3f5125daa9762..0000000000000000000000000000000000000000
--- a/ca-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:11723b88d86d6b94e7d0f7733beca5d7de1ab86b386ea66bc405cc1298b6f575
-size 164286
diff --git a/cs-en/test-00000-of-00001.parquet b/cs-en/test-00000-of-00001.parquet
deleted file mode 100644
index b4ba47eb58a00538ed5cc14afd17de7927ae1cac..0000000000000000000000000000000000000000
--- a/cs-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f6c1fe6df46dbd37d0a08dfea114b9b05d1d24ec6c081d2050f71577e0f2402e
-size 164899
diff --git a/cs-en/train-00000-of-00001.parquet b/cs-en/train-00000-of-00001.parquet
deleted file mode 100644
index ac025d8d40f5c3d2799298f323c0017994444818..0000000000000000000000000000000000000000
--- a/cs-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:35d86712cc06db477d813d040130d990deff9314f9785caa4e82c651cdebe9de
-size 72691523
diff --git a/cs-en/validation-00000-of-00001.parquet b/cs-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 7f93627ee3c27c7424abef19e2e992c92b3250c9..0000000000000000000000000000000000000000
--- a/cs-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:77ea5241063bff5da30f679dba706327c31a54e69d52a43b7285d0def471f0f1
-size 172092
diff --git a/cy-en/test-00000-of-00001.parquet b/cy-en/test-00000-of-00001.parquet
deleted file mode 100644
index e0c37c94ba6c16be54184be222640ed330c76d34..0000000000000000000000000000000000000000
--- a/cy-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7b4f8dc9af068cad44f3f010e32ad672329fa1a009784aadf3115614dacdff4b
-size 85616
diff --git a/cy-en/train-00000-of-00001.parquet b/cy-en/train-00000-of-00001.parquet
deleted file mode 100644
index 06887f4acdfc03a2c109acba6b92d495097fa3c9..0000000000000000000000000000000000000000
--- a/cy-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f81d27f41bd2607e472f8025a6b92da805842ddf648c4f7ddbb02223ca91da39
-size 13227579
diff --git a/cy-en/validation-00000-of-00001.parquet b/cy-en/validation-00000-of-00001.parquet
deleted file mode 100644
index d026a381b1bbf58aa8fd10c1e01932aa7fb028e7..0000000000000000000000000000000000000000
--- a/cy-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:61a25df5f3799c04ed339b99e936d67cff56aca6e12756bcd099310243bee95d
-size 85570
diff --git a/da-en/test-00000-of-00001.parquet b/da-en/test-00000-of-00001.parquet
deleted file mode 100644
index 2bb38b01425fe27d91d86344f955ebed361e880a..0000000000000000000000000000000000000000
--- a/da-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2328611e11fbf4a97220e43fe9dd8fd234689cf5f7938e2ad532668ff81e2c86
-size 214744
diff --git a/da-en/train-00000-of-00001.parquet b/da-en/train-00000-of-00001.parquet
deleted file mode 100644
index 516776b6a6bbfcdb516da4c2dc8631efc809eee1..0000000000000000000000000000000000000000
--- a/da-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:19e066c9cab308979df2d61c2bc2d6b9890a62e0e8201c080293b83bc7093859
-size 90573706
diff --git a/da-en/validation-00000-of-00001.parquet b/da-en/validation-00000-of-00001.parquet
deleted file mode 100644
index b52dcee77d3ddbd8a8eb222bad14ab5aa934df00..0000000000000000000000000000000000000000
--- a/da-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:47465ce170db040a9dc7a8f8ac6810e4ff806c5d6d0b375f35084128b224db1e
-size 216802
diff --git a/dataset_infos.json b/dataset_infos.json
new file mode 100644
index 0000000000000000000000000000000000000000..f0bb43e07e2054588c87d3c725de2533812a4067
--- /dev/null
+++ b/dataset_infos.json
@@ -0,0 +1 @@
+{"af-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["af", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "af", "output": "en"}, "builder_name": "opus100", "config_name": "af-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 135916, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 18726471, "num_examples": 275512, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 132777, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-af-en-v1.0.tar.gz": {"num_bytes": 7505036, "checksum": "d5b1ac19fdc41508fdfd946e1c51ae6777d181474f5e77b1fee88bf76d254d52"}}, "download_size": 7505036, "post_processing_size": null, "dataset_size": 18995164, "size_in_bytes": 26500200}, "am-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["am", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "am", "output": "en"}, "builder_name": "opus100", "config_name": "am-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 588029, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 21950644, "num_examples": 89027, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 566077, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-am-en-v1.0.tar.gz": {"num_bytes": 7004193, "checksum": "c68d7590fd4c708d6b80639389cf73bd1ebc3ca39bbaeaaaf7c6d48140c780c4"}}, "download_size": 7004193, "post_processing_size": null, "dataset_size": 23104750, "size_in_bytes": 30108943}, "an-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["an", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "an", "output": "en"}, "builder_name": "opus100", "config_name": "an-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 438332, "num_examples": 6961, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-an-en-v1.0.tar.gz": {"num_bytes": 96148, "checksum": "8083ade0f12af4abf1600f497f4327af45a4d074cf85134a82e07bd7dfdff876"}}, "download_size": 96148, "post_processing_size": null, "dataset_size": 438332, "size_in_bytes": 534480}, "ar-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "en"}, "builder_name": "opus100", "config_name": "ar-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 331648, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 152766484, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 2272106, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-ar-en-v1.0.tar.gz": {"num_bytes": 55286865, "checksum": "efc55eaab636e5c469b1ab3e2543bcfdc57f375ee6d57f248cfcfa589b46594b"}}, "download_size": 55286865, "post_processing_size": null, "dataset_size": 155370238, "size_in_bytes": 210657103}, "as-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["as", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "as", "output": "en"}, "builder_name": "opus100", "config_name": "as-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 261466, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 15634648, "num_examples": 138479, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 248139, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-as-en-v1.0.tar.gz": {"num_bytes": 4183517, "checksum": "adf38305e6256bb5e338b6c9d9dd08b5755acb637b255196dda1b78e23e130dd"}}, "download_size": 4183517, "post_processing_size": null, "dataset_size": 16144253, "size_in_bytes": 20327770}, "az-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["az", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "az", "output": "en"}, "builder_name": "opus100", "config_name": "az-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 393109, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 56431259, "num_examples": 262089, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 407109, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-az-en-v1.0.tar.gz": {"num_bytes": 18897341, "checksum": "a3a3ba50c98ffa8cfb453d0023be3da9f4e5c6cf03ce03ae3c402b1c8745c7ab"}}, "download_size": 18897341, "post_processing_size": null, "dataset_size": 57231477, "size_in_bytes": 76128818}, "be-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["be", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "be", "output": "en"}, "builder_name": "opus100", "config_name": "be-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 166858, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 5298500, "num_examples": 67312, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 175205, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-be-en-v1.0.tar.gz": {"num_bytes": 1906088, "checksum": "f5a047de898b222f413bea7316ef60fbad99a5f4ab11b3041e1ea0d9d4439dab"}}, "download_size": 1906088, "post_processing_size": null, "dataset_size": 5640563, "size_in_bytes": 7546651}, "bg-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["bg", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "bg", "output": "en"}, "builder_name": "opus100", "config_name": "bg-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 243751, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 108930347, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 234848, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-bg-en-v1.0.tar.gz": {"num_bytes": 36980744, "checksum": "ee019090235379a17e90422fe739667b3af913da1ee55e662db10db317d6ec5a"}}, "download_size": 36980744, "post_processing_size": null, "dataset_size": 109408946, "size_in_bytes": 146389690}, "bn-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["bn", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "bn", "output": "en"}, "builder_name": "opus100", "config_name": "bn-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 510101, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 249906846, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 498414, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-bn-en-v1.0.tar.gz": {"num_bytes": 72999655, "checksum": "03cecc49791aef3788ebe6bd36908462e3e198916d42ca1165c74c7443580845"}}, "download_size": 72999655, "post_processing_size": null, "dataset_size": 250915361, "size_in_bytes": 323915016}, "br-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["br", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "br", "output": "en"}, "builder_name": "opus100", "config_name": "br-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 127925, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 8539006, "num_examples": 153447, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 133772, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-br-en-v1.0.tar.gz": {"num_bytes": 3323458, "checksum": "1dde74dd2f935df7bb5db8909c77bd0e19cef97e60195c9e9c2154a8f270af08"}}, "download_size": 3323458, "post_processing_size": null, "dataset_size": 8800703, "size_in_bytes": 12124161}, "bs-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["bs", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "bs", "output": "en"}, "builder_name": "opus100", "config_name": "bs-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 168622, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 75082948, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 172481, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-bs-en-v1.0.tar.gz": {"num_bytes": 30746956, "checksum": "de8747c8b5dd34f5e19aa9ccf3734626cf7f7c0b82ef4920a33e3c722acb70e2"}}, "download_size": 30746956, "post_processing_size": null, "dataset_size": 75424051, "size_in_bytes": 106171007}, "ca-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ca", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ca", "output": "en"}, "builder_name": "opus100", "config_name": "ca-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 205666, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 88405510, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 212637, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-ca-en-v1.0.tar.gz": {"num_bytes": 36267794, "checksum": "40aa2a23d4fb5408b05f9a2f95544e7b7195f7e0aa405fac93263930a77ab19a"}}, "download_size": 36267794, "post_processing_size": null, "dataset_size": 88823813, "size_in_bytes": 125091607}, "cs-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["cs", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "cs", "output": "en"}, "builder_name": "opus100", "config_name": "cs-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 205274, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 91897719, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 219084, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-cs-en-v1.0.tar.gz": {"num_bytes": 39673827, "checksum": "e870a377d0eb7327ad9f5e4d979feb41ad0e0a331bc176fcdd7e45ecdda1ff0f"}}, "download_size": 39673827, "post_processing_size": null, "dataset_size": 92322077, "size_in_bytes": 131995904}, "cy-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["cy", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "cy", "output": "en"}, "builder_name": "opus100", "config_name": "cy-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 124289, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 17244980, "num_examples": 289521, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 118856, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-cy-en-v1.0.tar.gz": {"num_bytes": 6487005, "checksum": "935145550ed2fbef59ba947b9a15d5281039fbbc5377624af3011ed4d454f6b0"}}, "download_size": 6487005, "post_processing_size": null, "dataset_size": 17488125, "size_in_bytes": 23975130}, "da-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["da", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "da", "output": "en"}, "builder_name": "opus100", "config_name": "da-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 298123, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 126425274, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 300624, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-da-en-v1.0.tar.gz": {"num_bytes": 50404122, "checksum": "20fcc66a488897a549495a11427a9f3b84926a323ddaa0372f03b14fabee4f14"}}, "download_size": 50404122, "post_processing_size": null, "dataset_size": 127024021, "size_in_bytes": 177428143}, "de-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["de", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "de", "output": "en"}, "builder_name": "opus100", "config_name": "de-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 330959, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 152246756, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 332350, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-de-en-v1.0.tar.gz": {"num_bytes": 67205361, "checksum": "aca2cb2f71268700c5734c7b177fbc91dcadc1ac59844000a85d9d10d0bb19c0"}}, "download_size": 67205361, "post_processing_size": null, "dataset_size": 152910065, "size_in_bytes": 220115426}, "dz-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["dz", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "dz", "output": "en"}, "builder_name": "opus100", "config_name": "dz-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 81162, "num_examples": 624, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-dz-en-v1.0.tar.gz": {"num_bytes": 17814, "checksum": "c873dfd45a93420ac96bdd05d2835a7ee57bceff7f2258cfda8aecb2787bbe05"}}, "download_size": 17814, "post_processing_size": null, "dataset_size": 81162, "size_in_bytes": 98976}, "el-en": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["el", "en"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "el", "output": "en"}, "builder_name": "opus100", "config_name": "el-en", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 302393, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 127964703, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 291234, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-el-en-v1.0.tar.gz": {"num_bytes": 43973686, "checksum": "2992dd7a8cadd2c3db7b7267dd3df4cc5f3dee21a2514d6218a718891f543f99"}}, "download_size": 43973686, "post_processing_size": null, "dataset_size": 128558330, "size_in_bytes": 172532016}, "en-eo": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "eo"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "eo"}, "builder_name": "opus100", "config_name": "en-eo", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 167386, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 24431953, "num_examples": 337106, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 168838, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-eo-v1.0.tar.gz": {"num_bytes": 9999313, "checksum": "1db2518821cac8d313085f2a79dc3dcaa30ca35dab8436786e70921209c52a87"}}, "download_size": 9999313, "post_processing_size": null, "dataset_size": 24768177, "size_in_bytes": 34767490}, "en-es": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "es"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "es"}, "builder_name": "opus100", "config_name": "en-es", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 326270, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 136643904, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 326735, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-es-v1.0.tar.gz": {"num_bytes": 55534068, "checksum": "e361cd112e9831d50e48d59fa0a9630a51daf323ed6e555fa25464b5e409c649"}}, "download_size": 55534068, "post_processing_size": null, "dataset_size": 137296909, "size_in_bytes": 192830977}, "en-et": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "et"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "et"}, "builder_name": "opus100", "config_name": "en-et", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 272171, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 112299053, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 276962, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-et-v1.0.tar.gz": {"num_bytes": 46235623, "checksum": "9fefd0239214e3620b8e525c7bab5ddbac4b8cdcdaf784936be9cedde5cef421"}}, "download_size": 46235623, "post_processing_size": null, "dataset_size": 112848186, "size_in_bytes": 159083809}, "en-eu": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "eu"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "eu"}, "builder_name": "opus100", "config_name": "en-eu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 280885, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 112330085, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 281503, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-eu-v1.0.tar.gz": {"num_bytes": 46389313, "checksum": "b280ab292fad2b0c20599b0c0032eb1232706327228cdfe91611bbcf64979b9a"}}, "download_size": 46389313, "post_processing_size": null, "dataset_size": 112892473, "size_in_bytes": 159281786}, "en-fa": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "fa"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "fa"}, "builder_name": "opus100", "config_name": "en-fa", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 296556, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 125401335, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 291129, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-fa-v1.0.tar.gz": {"num_bytes": 44568447, "checksum": "000305cca8a8c6278f8d036c06845f4c994d1343f4710873f99b705aeb529e51"}}, "download_size": 44568447, "post_processing_size": null, "dataset_size": 125989020, "size_in_bytes": 170557467}, "en-fi": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "fi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "fi"}, "builder_name": "opus100", "config_name": "en-fi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 245822, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 106025790, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 247227, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-fi-v1.0.tar.gz": {"num_bytes": 42563103, "checksum": "fa888ce283e2887a6932ab007518f2b2586e6d80ebe44856f8648e47e0c83b0f"}}, "download_size": 42563103, "post_processing_size": null, "dataset_size": 106518839, "size_in_bytes": 149081942}, "en-fr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "fr"}, "builder_name": "opus100", "config_name": "en-fr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 469731, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 201441250, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 481484, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-fr-v1.0.tar.gz": {"num_bytes": 81009778, "checksum": "c5521d61d7a158bcdebdd590c86b6e26e7d434cb1e02ca12681f8f82ed998c2d"}}, "download_size": 81009778, "post_processing_size": null, "dataset_size": 202392465, "size_in_bytes": 283402243}, "en-fy": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "fy"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "fy"}, "builder_name": "opus100", "config_name": "en-fy", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 101246, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 3895688, "num_examples": 54342, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 100129, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-fy-v1.0.tar.gz": {"num_bytes": 1522187, "checksum": "bc86f0ee0e99d4fd9e2bb3c1c27ce8f50ee0ca56fe12b75e3d8d4abb97a267eb"}}, "download_size": 1522187, "post_processing_size": null, "dataset_size": 4097063, "size_in_bytes": 5619250}, "en-ga": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ga"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ga"}, "builder_name": "opus100", "config_name": "en-ga", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 503317, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 42132742, "num_examples": 289524, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 503217, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ga-v1.0.tar.gz": {"num_bytes": 14998873, "checksum": "8f278d5c9cdc9e4ce660a9afe72a286ef4651adfb9f4d2fe0f38b4cacb8be449"}}, "download_size": 14998873, "post_processing_size": null, "dataset_size": 43139276, "size_in_bytes": 58138149}, "en-gd": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "gd"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "gd"}, "builder_name": "opus100", "config_name": "en-gd", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 218362, "num_examples": 1606, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1254795, "num_examples": 16316, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 203885, "num_examples": 1605, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-gd-v1.0.tar.gz": {"num_bytes": 564053, "checksum": "fc567fb3065c700046120bbdcd3b5a54eb8f2cfe3d7a97dcad5270ff2e342343"}}, "download_size": 564053, "post_processing_size": null, "dataset_size": 1677042, "size_in_bytes": 2241095}, "en-gl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "gl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "gl"}, "builder_name": "opus100", "config_name": "en-gl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 190699, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 43327444, "num_examples": 515344, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 193606, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-gl-v1.0.tar.gz": {"num_bytes": 18056665, "checksum": "ff2e4ca48fd491a8bff5400a5584087ba05c893778845f2167c218ad2d57f4a3"}}, "download_size": 18056665, "post_processing_size": null, "dataset_size": 43711749, "size_in_bytes": 61768414}, "en-gu": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "gu"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "gu"}, "builder_name": "opus100", "config_name": "en-gu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 199733, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 33641975, "num_examples": 318306, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 205550, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-gu-v1.0.tar.gz": {"num_bytes": 9407543, "checksum": "ea614ac05ca527f903604603cf674b2825058e12a1bfb74eba40c77d52e773f5"}}, "download_size": 9407543, "post_processing_size": null, "dataset_size": 34047258, "size_in_bytes": 43454801}, "en-ha": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ha"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ha"}, "builder_name": "opus100", "config_name": "en-ha", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 407352, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 20391964, "num_examples": 97983, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 411526, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ha-v1.0.tar.gz": {"num_bytes": 6898482, "checksum": "e2c24892c72a530e43eaea0e3033291b8ea166a2c8665955a074f56cacd69136"}}, "download_size": 6898482, "post_processing_size": null, "dataset_size": 21210842, "size_in_bytes": 28109324}, "en-he": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "he"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "he"}, "builder_name": "opus100", "config_name": "en-he", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 208475, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 91160431, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 209446, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-he-v1.0.tar.gz": {"num_bytes": 31214136, "checksum": "693831c44c3afc5f34d7e3881dddd9eea6ba9ad5ec08d63359f4ab9bb033a758"}}, "download_size": 31214136, "post_processing_size": null, "dataset_size": 91578352, "size_in_bytes": 122792488}, "en-hi": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "hi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "hi"}, "builder_name": "opus100", "config_name": "en-hi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 496578, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 124923977, "num_examples": 534319, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 474087, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-hi-v1.0.tar.gz": {"num_bytes": 35993452, "checksum": "f6efc00fd95972b9b13157167356152766d962db07b6e73e6cc87479083c3b90"}}, "download_size": 35993452, "post_processing_size": null, "dataset_size": 125894642, "size_in_bytes": 161888094}, "en-hr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "hr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "hr"}, "builder_name": "opus100", "config_name": "en-hr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 179644, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 75310316, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 179623, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-hr-v1.0.tar.gz": {"num_bytes": 30728154, "checksum": "7b2302e36e747700ed0036c119004d31a7955918af7e1f11dc868e81c4c9f537"}}, "download_size": 30728154, "post_processing_size": null, "dataset_size": 75669583, "size_in_bytes": 106397737}, "en-hu": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "hu"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "hu"}, "builder_name": "opus100", "config_name": "en-hu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 206047, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 87484262, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 208315, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-hu-v1.0.tar.gz": {"num_bytes": 35696235, "checksum": "04b51fd88964c0a42f49dc779e1d515944872ae1a998a6d3602cf63aa4d2d216"}}, "download_size": 35696235, "post_processing_size": null, "dataset_size": 87898624, "size_in_bytes": 123594859}, "en-hy": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "hy"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "hy"}, "builder_name": "opus100", "config_name": "en-hy", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 652631, "num_examples": 7059, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-hy-v1.0.tar.gz": {"num_bytes": 215246, "checksum": "dbcd452a3a0eb3c1485621ea314a096ab5ddf89e84e1f02fe7b4bdf4cc268dcc"}}, "download_size": 215246, "post_processing_size": null, "dataset_size": 652631, "size_in_bytes": 867877}, "en-id": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "id"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "id"}, "builder_name": "opus100", "config_name": "en-id", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 177693, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 78699773, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 180032, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-id-v1.0.tar.gz": {"num_bytes": 29914089, "checksum": "3466dd5f90b0e44542edafac932c25fbeabcad65b5cd50d264319f1fd18c1b8d"}}, "download_size": 29914089, "post_processing_size": null, "dataset_size": 79057498, "size_in_bytes": 108971587}, "en-ig": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ig"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ig"}, "builder_name": "opus100", "config_name": "en-ig", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 137332, "num_examples": 1843, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1612539, "num_examples": 18415, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 135995, "num_examples": 1843, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ig-v1.0.tar.gz": {"num_bytes": 391849, "checksum": "4cbee7925658026f1791d2638edbc1f72435306f39428f76627fa84506142755"}}, "download_size": 391849, "post_processing_size": null, "dataset_size": 1885866, "size_in_bytes": 2277715}, "en-is": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "is"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "is"}, "builder_name": "opus100", "config_name": "en-is", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 170887, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 73964915, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 170640, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-is-v1.0.tar.gz": {"num_bytes": 28831218, "checksum": "1dfc59b45631dbf849c75a23b14379b05296cf651e24dd2df7ab3ab85f916624"}}, "download_size": 28831218, "post_processing_size": null, "dataset_size": 74306442, "size_in_bytes": 103137660}, "en-it": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "it"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "it"}, "builder_name": "opus100", "config_name": "en-it", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 299037, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 123655086, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 294362, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-it-v1.0.tar.gz": {"num_bytes": 50903618, "checksum": "54da7db3f7bcbd21e8b41180551d4fd5de0eb1802bfb2ec2b5d44bb919ba464a"}}, "download_size": 50903618, "post_processing_size": null, "dataset_size": 124248485, "size_in_bytes": 175152103}, "en-ja": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ja"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ja"}, "builder_name": "opus100", "config_name": "en-ja", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 190999, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 88349369, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 191419, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ja-v1.0.tar.gz": {"num_bytes": 34452575, "checksum": "78e83ee1b0e17834a02c2f51df1ce71d455b9211e46f113b97d441e40fed0681"}}, "download_size": 34452575, "post_processing_size": null, "dataset_size": 88731787, "size_in_bytes": 123184362}, "en-ka": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ka"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ka"}, "builder_name": "opus100", "config_name": "en-ka", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 256227, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 42465706, "num_examples": 377306, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 260416, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ka-v1.0.tar.gz": {"num_bytes": 12743188, "checksum": "6b189050fad96016e21a24869b8164da7450214197c473584ed1445bf626f617"}}, "download_size": 12743188, "post_processing_size": null, "dataset_size": 42982349, "size_in_bytes": 55725537}, "en-kk": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "kk"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "kk"}, "builder_name": "opus100", "config_name": "en-kk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 137664, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 7124378, "num_examples": 79927, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 139665, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-kk-v1.0.tar.gz": {"num_bytes": 2425372, "checksum": "be76b8a0d8a60809cbb90320e7d484501c86686eb65c5aac585ad94b979da850"}}, "download_size": 2425372, "post_processing_size": null, "dataset_size": 7401707, "size_in_bytes": 9827079}, "en-km": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "km"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "km"}, "builder_name": "opus100", "config_name": "en-km", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 289027, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 19680611, "num_examples": 111483, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 302527, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-km-v1.0.tar.gz": {"num_bytes": 5193620, "checksum": "b65cbbf241a1eeeafd6b39c366f219356e6ce6e065d146a1e847200f055dd39b"}}, "download_size": 5193620, "post_processing_size": null, "dataset_size": 20272165, "size_in_bytes": 25465785}, "en-ko": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ko"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ko"}, "builder_name": "opus100", "config_name": "en-ko", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 190696, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 93665332, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 189368, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ko-v1.0.tar.gz": {"num_bytes": 37602794, "checksum": "f9040eb06807cf1cb2c0d3352249b38546bdbff668c242b98c95bfae84190fd2"}}, "download_size": 37602794, "post_processing_size": null, "dataset_size": 94045396, "size_in_bytes": 131648190}, "en-kn": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "kn"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "kn"}, "builder_name": "opus100", "config_name": "en-kn", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 77205, "num_examples": 918, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1833334, "num_examples": 14537, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 77607, "num_examples": 917, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-kn-v1.0.tar.gz": {"num_bytes": 525449, "checksum": "5c100de50142e36e750e6a0d595aeedccda7f1479d781b4a4a337c8361bfe87a"}}, "download_size": 525449, "post_processing_size": null, "dataset_size": 1988146, "size_in_bytes": 2513595}, "en-ku": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ku"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ku"}, "builder_name": "opus100", "config_name": "en-ku", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 247847, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 49107864, "num_examples": 144844, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 239325, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ku-v1.0.tar.gz": {"num_bytes": 14252198, "checksum": "b198288af58893bcb19e999565eb0180d100505003d5064643f51329024bf117"}}, "download_size": 14252198, "post_processing_size": null, "dataset_size": 49595036, "size_in_bytes": 63847234}, "en-ky": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ky"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ky"}, "builder_name": "opus100", "config_name": "en-ky", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 142530, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1879298, "num_examples": 27215, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 138487, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ky-v1.0.tar.gz": {"num_bytes": 616902, "checksum": "2694ed66518822f283bfa8097e94160256e1dc9c405545258eb7962282ec62b4"}}, "download_size": 616902, "post_processing_size": null, "dataset_size": 2160315, "size_in_bytes": 2777217}, "en-li": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "li"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "li"}, "builder_name": "opus100", "config_name": "en-li", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 93350, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1628601, "num_examples": 25535, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 92906, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-li-v1.0.tar.gz": {"num_bytes": 450092, "checksum": "e17e96b9331e90542e3a5145ae27e36a29a6433cbe1e4b6352bd6634546dfe7c"}}, "download_size": 450092, "post_processing_size": null, "dataset_size": 1814857, "size_in_bytes": 2264949}, "en-lt": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "lt"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "lt"}, "builder_name": "opus100", "config_name": "en-lt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 482615, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 177061044, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 469117, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-lt-v1.0.tar.gz": {"num_bytes": 69388131, "checksum": "af798e0d92f37805bc9a8f3a13ff48a1e016b67938062016558f3a7cc39f8ded"}}, "download_size": 69388131, "post_processing_size": null, "dataset_size": 178012776, "size_in_bytes": 247400907}, "en-lv": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "lv"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "lv"}, "builder_name": "opus100", "config_name": "en-lv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 536576, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 206051849, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 522072, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-lv-v1.0.tar.gz": {"num_bytes": 78952903, "checksum": "5b824a15347a139655a8e630a724bc6df2d7d454d35b20da028630aa4969bb53"}}, "download_size": 78952903, "post_processing_size": null, "dataset_size": 207110497, "size_in_bytes": 286063400}, "en-mg": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "mg"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "mg"}, "builder_name": "opus100", "config_name": "en-mg", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 525067, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 130865649, "num_examples": 590771, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 511171, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-mg-v1.0.tar.gz": {"num_bytes": 52470504, "checksum": "b6f7242fd0d535422ca8958fdf295808ef5a47bb3f9901d26dc08ff2a8cbb843"}}, "download_size": 52470504, "post_processing_size": null, "dataset_size": 131901887, "size_in_bytes": 184372391}, "en-mk": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "mk"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "mk"}, "builder_name": "opus100", "config_name": "en-mk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 308934, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 117069489, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 305498, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-mk-v1.0.tar.gz": {"num_bytes": 39517761, "checksum": "2fdb6bbfb189dbd6db5cea247488b26bf0d943cb82415a5f4cbe2d664d964fb1"}}, "download_size": 39517761, "post_processing_size": null, "dataset_size": 117683921, "size_in_bytes": 157201682}, "en-ml": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ml"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ml"}, "builder_name": "opus100", "config_name": "en-ml", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 340626, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 199971743, "num_examples": 822746, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 334459, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ml-v1.0.tar.gz": {"num_bytes": 48654808, "checksum": "0121284cdb5582b95aa344de6fd31322202da58e938763037e5cbf40d39e7644"}}, "download_size": 48654808, "post_processing_size": null, "dataset_size": 200646828, "size_in_bytes": 249301636}, "en-mn": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "mn"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "mn"}, "builder_name": "opus100", "config_name": "en-mn", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 250778, "num_examples": 4294, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-mn-v1.0.tar.gz": {"num_bytes": 42039, "checksum": "52cbcf55978c5cc6be6062a3669edeb0c4833e0e8cdc235295a6c011b4fdaf7b"}}, "download_size": 42039, "post_processing_size": null, "dataset_size": 250778, "size_in_bytes": 292817}, "en-mr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "mr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "mr"}, "builder_name": "opus100", "config_name": "en-mr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 238612, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 2724131, "num_examples": 27007, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 235540, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-mr-v1.0.tar.gz": {"num_bytes": 910211, "checksum": "0988cfa9bac2da79b15c2a7a1cde0e31851b0d418c1edf3a0027c9b6fa6d43c2"}}, "download_size": 910211, "post_processing_size": null, "dataset_size": 3198283, "size_in_bytes": 4108494}, "en-ms": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ms"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ms"}, "builder_name": "opus100", "config_name": "en-ms", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 179705, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 76829645, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 180183, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ms-v1.0.tar.gz": {"num_bytes": 29807607, "checksum": "9f46a5e7dee938b2f9e51182b0d7939a3cea61c5cdda6d4fc0ba724d99688cd4"}}, "download_size": 29807607, "post_processing_size": null, "dataset_size": 77189533, "size_in_bytes": 106997140}, "en-mt": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "mt"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "mt"}, "builder_name": "opus100", "config_name": "en-mt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 566134, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 222222396, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 594386, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-mt-v1.0.tar.gz": {"num_bytes": 84757608, "checksum": "327f2fbab86f4aec9df76c7e39d3f32b8f56b97877df1579f73c10f431f7305e"}}, "download_size": 84757608, "post_processing_size": null, "dataset_size": 223382916, "size_in_bytes": 308140524}, "en-my": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "my"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "my"}, "builder_name": "opus100", "config_name": "en-my", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 337351, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 3673501, "num_examples": 24594, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 336155, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-my-v1.0.tar.gz": {"num_bytes": 1038600, "checksum": "5338dfdbdea5a6178f4b3654e4f8644906d7c6d9b102d7e9c3631b65ab9debc9"}}, "download_size": 1038600, "post_processing_size": null, "dataset_size": 4347007, "size_in_bytes": 5385607}, "en-nb": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "nb"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "nb"}, "builder_name": "opus100", "config_name": "en-nb", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 334117, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 13611709, "num_examples": 142906, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 324400, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-nb-v1.0.tar.gz": {"num_bytes": 5706626, "checksum": "f184509d7a781fe60aaf617ca23a373488c369ca8eb6d5c2eebf7c74e747d6a7"}}, "download_size": 5706626, "post_processing_size": null, "dataset_size": 14270226, "size_in_bytes": 19976852}, "en-ne": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ne"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ne"}, "builder_name": "opus100", "config_name": "en-ne", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 186527, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 44136280, "num_examples": 406381, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 204920, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ne-v1.0.tar.gz": {"num_bytes": 11711988, "checksum": "0939f62d42db0f070c309b0fb02c4733e3e943cdbc504739fc0e4b42cbcd3ba8"}}, "download_size": 11711988, "post_processing_size": null, "dataset_size": 44527727, "size_in_bytes": 56239715}, "en-nl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "nl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "nl"}, "builder_name": "opus100", "config_name": "en-nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 282755, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 112327073, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 270940, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-nl-v1.0.tar.gz": {"num_bytes": 45374708, "checksum": "8d59b586929585de9aa0f9805bc9873c5b9ea5075d34bf23ffb2714b02499308"}}, "download_size": 45374708, "post_processing_size": null, "dataset_size": 112880768, "size_in_bytes": 158255476}, "en-nn": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "nn"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "nn"}, "builder_name": "opus100", "config_name": "en-nn", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 179007, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 32924821, "num_examples": 486055, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 187650, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-nn-v1.0.tar.gz": {"num_bytes": 12742134, "checksum": "31c1fde5fb4b95c37cd9ceafcaf6104574727dc31b15d4b255f0ab71df4c6579"}}, "download_size": 12742134, "post_processing_size": null, "dataset_size": 33291478, "size_in_bytes": 46033612}, "en-no": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "no"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "no"}, "builder_name": "opus100", "config_name": "en-no", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 173328, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 74106283, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 178013, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-no-v1.0.tar.gz": {"num_bytes": 28851262, "checksum": "a8b20ca8dfd25581b54303a780c19105eab3645ad7981a1a821e93d634bfc4f6"}}, "download_size": 28851262, "post_processing_size": null, "dataset_size": 74457624, "size_in_bytes": 103308886}, "en-oc": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "oc"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "oc"}, "builder_name": "opus100", "config_name": "en-oc", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 82350, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1627206, "num_examples": 35791, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 81650, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-oc-v1.0.tar.gz": {"num_bytes": 607192, "checksum": "9fe531860f1a7f89d97053fc4144b71fdbf6e734dc2d3d52dbc20714493cd0d4"}}, "download_size": 607192, "post_processing_size": null, "dataset_size": 1791206, "size_in_bytes": 2398398}, "en-or": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "or"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "or"}, "builder_name": "opus100", "config_name": "en-or", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 163947, "num_examples": 1318, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1500749, "num_examples": 14273, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 155331, "num_examples": 1317, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-or-v1.0.tar.gz": {"num_bytes": 499401, "checksum": "dc15a00dbb19541ae46491490184d08302095faf7ed88d962f45cec768d87b05"}}, "download_size": 499401, "post_processing_size": null, "dataset_size": 1820027, "size_in_bytes": 2319428}, "en-pa": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "pa"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "pa"}, "builder_name": "opus100", "config_name": "en-pa", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 133909, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 8509228, "num_examples": 107296, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 136196, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-pa-v1.0.tar.gz": {"num_bytes": 2589682, "checksum": "1107d5ae7c5b5c6a8575bce896cecb1e2b45b7dc454294222062245b6fe3f947"}}, "download_size": 2589682, "post_processing_size": null, "dataset_size": 8779333, "size_in_bytes": 11369015}, "en-pl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "pl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "pl"}, "builder_name": "opus100", "config_name": "en-pl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 212503, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 95248523, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 218216, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-pl-v1.0.tar.gz": {"num_bytes": 39320454, "checksum": "8684ef187956bfab33896f84437722e2c41ea00341eee7096039f0e39c4f2ddc"}}, "download_size": 39320454, "post_processing_size": null, "dataset_size": 95679242, "size_in_bytes": 134999696}, "en-ps": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ps"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ps"}, "builder_name": "opus100", "config_name": "en-ps", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 93003, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 4436576, "num_examples": 79127, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 95164, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ps-v1.0.tar.gz": {"num_bytes": 1223087, "checksum": "591b83f63a43b534653523681ac880bcaea2ff69c8f266112c66223bf44e90f5"}}, "download_size": 1223087, "post_processing_size": null, "dataset_size": 4624743, "size_in_bytes": 5847830}, "en-pt": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "pt"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "pt"}, "builder_name": "opus100", "config_name": "en-pt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 296122, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 118243649, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 292082, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-pt-v1.0.tar.gz": {"num_bytes": 48087550, "checksum": "550e3382309dee492c77f8ce88c32fb559211ad4f032dca2d34c10ec0b0fe75d"}}, "download_size": 48087550, "post_processing_size": null, "dataset_size": 118831853, "size_in_bytes": 166919403}, "en-ro": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ro"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ro"}, "builder_name": "opus100", "config_name": "en-ro", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 198647, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 85249851, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 199172, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ro-v1.0.tar.gz": {"num_bytes": 35032743, "checksum": "4c87e36f2ff5eae5ffefb66abcf6c65a5f8af673946c911f10ab42bc468c5d42"}}, "download_size": 35032743, "post_processing_size": null, "dataset_size": 85647670, "size_in_bytes": 120680413}, "en-ru": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ru"}, "builder_name": "opus100", "config_name": "en-ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 490984, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 195101737, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 490246, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ru-v1.0.tar.gz": {"num_bytes": 68501634, "checksum": "422c96c952e37a0e7ae924538f6c521a9dcedb200e0b29eab07a6e3fd00cca9d"}}, "download_size": 68501634, "post_processing_size": null, "dataset_size": 196082967, "size_in_bytes": 264584601}, "en-rw": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "rw"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "rw"}, "builder_name": "opus100", "config_name": "en-rw", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 136197, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 15286303, "num_examples": 173823, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 134965, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-rw-v1.0.tar.gz": {"num_bytes": 5233241, "checksum": "ac65b6ee8d4900c078a53d90af463952b1f2f67333f36ccd53bc34cafa424ee4"}}, "download_size": 5233241, "post_processing_size": null, "dataset_size": 15557465, "size_in_bytes": 20790706}, "en-se": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "se"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "se"}, "builder_name": "opus100", "config_name": "en-se", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 85705, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 2047412, "num_examples": 35907, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 83672, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-se-v1.0.tar.gz": {"num_bytes": 806982, "checksum": "ccc109d695c8d238f5d8c318b1a329b4f4bd434452bd178c0db90e7df60bae0a"}}, "download_size": 806982, "post_processing_size": null, "dataset_size": 2216789, "size_in_bytes": 3023771}, "en-sh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sh"}, "builder_name": "opus100", "config_name": "en-sh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 569487, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 60900239, "num_examples": 267211, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 555602, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sh-v1.0.tar.gz": {"num_bytes": 22357505, "checksum": "68018403e648b55cca750e13d7c91cf6ec12be3a3cabcf83247200150fa39594"}}, "download_size": 22357505, "post_processing_size": null, "dataset_size": 62025328, "size_in_bytes": 84382833}, "en-si": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "si"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "si"}, "builder_name": "opus100", "config_name": "en-si", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 271743, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 114951675, "num_examples": 979109, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 271244, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-si-v1.0.tar.gz": {"num_bytes": 33247484, "checksum": "2a654e11df7040ab70f36cde1d84a431696cb834c5d25abe0bfbd375615262a6"}}, "download_size": 33247484, "post_processing_size": null, "dataset_size": 115494662, "size_in_bytes": 148742146}, "en-sk": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sk"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sk"}, "builder_name": "opus100", "config_name": "en-sk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 258042, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 111743868, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 255470, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sk-v1.0.tar.gz": {"num_bytes": 46618395, "checksum": "1641ebabc596e1dad84ec1bfccfdc0d8c65837fc8739631eb7826066bb328f05"}}, "download_size": 46618395, "post_processing_size": null, "dataset_size": 112257380, "size_in_bytes": 158875775}, "en-sl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sl"}, "builder_name": "opus100", "config_name": "en-sl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 205478, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 90270957, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 198662, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sl-v1.0.tar.gz": {"num_bytes": 37536724, "checksum": "57e5149d179d95eb7364220bcf10afbf419d94b402ad6dd8108610ae3e9e6760"}}, "download_size": 37536724, "post_processing_size": null, "dataset_size": 90675097, "size_in_bytes": 128211821}, "en-sq": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sq"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sq"}, "builder_name": "opus100", "config_name": "en-sq", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 275379, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 105745981, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 267312, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sq-v1.0.tar.gz": {"num_bytes": 42697338, "checksum": "e2a9b220ad63824188297d184596a449c4a8c1ef3e1a681b63d46a9819c9ba62"}}, "download_size": 42697338, "post_processing_size": null, "dataset_size": 106288672, "size_in_bytes": 148986010}, "en-sr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sr"}, "builder_name": "opus100", "config_name": "en-sr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 180232, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 75726835, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 184246, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sr-v1.0.tar.gz": {"num_bytes": 31260575, "checksum": "a5660570f8dbb47f1c275bce55542f0e1e1a938645d24c7bdd01c941d5481b4f"}}, "download_size": 31260575, "post_processing_size": null, "dataset_size": 76091313, "size_in_bytes": 107351888}, "en-sv": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "sv"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "sv"}, "builder_name": "opus100", "config_name": "en-sv", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 271014, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 116985953, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 279994, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-sv-v1.0.tar.gz": {"num_bytes": 46694960, "checksum": "ab83d088b08bc1a9f7da7895b4caf4dd467a57b1be9096bc0b5c88d045156119"}}, "download_size": 46694960, "post_processing_size": null, "dataset_size": 117536961, "size_in_bytes": 164231921}, "en-ta": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ta"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ta"}, "builder_name": "opus100", "config_name": "en-ta", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 351990, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 74044524, "num_examples": 227014, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 335557, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ta-v1.0.tar.gz": {"num_bytes": 17652443, "checksum": "612322d09aa48b7875ee5eee94791aa5bc22ebd2b0411997d66682a0329b61d4"}}, "download_size": 17652443, "post_processing_size": null, "dataset_size": 74732071, "size_in_bytes": 92384514}, "en-te": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "te"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "te"}, "builder_name": "opus100", "config_name": "en-te", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 190595, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 6688625, "num_examples": 64352, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 193666, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-te-v1.0.tar.gz": {"num_bytes": 2011832, "checksum": "8bb0c2405cc62b37c1ec21584765e703b5739c2730c04898e6a18ff416c08529"}}, "download_size": 2011832, "post_processing_size": null, "dataset_size": 7072886, "size_in_bytes": 9084718}, "en-tg": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "tg"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "tg"}, "builder_name": "opus100", "config_name": "en-tg", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 372120, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 35477177, "num_examples": 193882, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 371728, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-tg-v1.0.tar.gz": {"num_bytes": 11389877, "checksum": "30caf792aa42db984dcc7704a09250b15a1a1b46a7d82e79d2c116ee4c7eeefa"}}, "download_size": 11389877, "post_processing_size": null, "dataset_size": 36221025, "size_in_bytes": 47610902}, "en-th": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "th"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "th"}, "builder_name": "opus100", "config_name": "en-th", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 290581, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 132821031, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 288366, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-th-v1.0.tar.gz": {"num_bytes": 38147204, "checksum": "223dc991bd45ff25e96dada988f97bedb0c6f28e930bd8ab3ff1344e4a2a5bb6"}}, "download_size": 38147204, "post_processing_size": null, "dataset_size": 133399978, "size_in_bytes": 171547182}, "en-tk": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "tk"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "tk"}, "builder_name": "opus100", "config_name": "en-tk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 83886, "num_examples": 1852, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 719633, "num_examples": 13110, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 81014, "num_examples": 1852, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-tk-v1.0.tar.gz": {"num_bytes": 157481, "checksum": "afbbf17a9600d41176c522a760a54d1e143e7855063893934159cbda8e4e427b"}}, "download_size": 157481, "post_processing_size": null, "dataset_size": 884533, "size_in_bytes": 1042014}, "en-tr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "tr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "tr"}, "builder_name": "opus100", "config_name": "en-tr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 183833, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 78946365, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 181917, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-tr-v1.0.tar.gz": {"num_bytes": 30892429, "checksum": "0d4a941721721c94e99013052491894062d7782efb94a6593a91fd1f0f06cba0"}}, "download_size": 30892429, "post_processing_size": null, "dataset_size": 79312115, "size_in_bytes": 110204544}, "en-tt": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "tt"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "tt"}, "builder_name": "opus100", "config_name": "en-tt", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 693276, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 35313258, "num_examples": 100843, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 701670, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-tt-v1.0.tar.gz": {"num_bytes": 9940523, "checksum": "d23d7043364d55f5cf47abf93193a831bfdc5c0ec38052ac84bdb166b98348fb"}}, "download_size": 9940523, "post_processing_size": null, "dataset_size": 36708204, "size_in_bytes": 46648727}, "en-ug": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ug"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ug"}, "builder_name": "opus100", "config_name": "en-ug", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 620881, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 31576580, "num_examples": 72170, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 631236, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ug-v1.0.tar.gz": {"num_bytes": 8687743, "checksum": "176ccf6f66439cda78892db6c0f8bad67fc3f9ebe25feac7cd405ecff2d785d7"}}, "download_size": 8687743, "post_processing_size": null, "dataset_size": 32828697, "size_in_bytes": 41516440}, "en-uk": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "uk"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "uk"}, "builder_name": "opus100", "config_name": "en-uk", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 249750, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 104230356, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 247131, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-uk-v1.0.tar.gz": {"num_bytes": 37415496, "checksum": "09a689c33567da03908d134d240c6e353319408ada4fddd949e46bc7d2a675ea"}}, "download_size": 37415496, "post_processing_size": null, "dataset_size": 104727237, "size_in_bytes": 142142733}, "en-ur": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "ur"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "ur"}, "builder_name": "opus100", "config_name": "en-ur", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 538564, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 268961304, "num_examples": 753913, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 529316, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-ur-v1.0.tar.gz": {"num_bytes": 81092186, "checksum": "5e6dfe7bb1c956b37594a2b81213dde5f7dba1c0e49ecccdebcb4aa73af23711"}}, "download_size": 81092186, "post_processing_size": null, "dataset_size": 270029184, "size_in_bytes": 351121370}, "en-uz": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "uz"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "uz"}, "builder_name": "opus100", "config_name": "en-uz", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 408683, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 38375434, "num_examples": 173157, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 398861, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-uz-v1.0.tar.gz": {"num_bytes": 11791643, "checksum": "364902a0e71c5e7b3dcb6bae4517fd094a445cad02566d81889222b06bed2650"}}, "download_size": 11791643, "post_processing_size": null, "dataset_size": 39182978, "size_in_bytes": 50974621}, "en-vi": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "vi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "vi"}, "builder_name": "opus100", "config_name": "en-vi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 192752, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 82615270, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 194729, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-vi-v1.0.tar.gz": {"num_bytes": 30647296, "checksum": "6a37b14536cfa554ea90f8560e3f93e5273c5783e062fc07bb2e3ffbc07daf95"}}, "download_size": 30647296, "post_processing_size": null, "dataset_size": 83002751, "size_in_bytes": 113650047}, "en-wa": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "wa"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "wa"}, "builder_name": "opus100", "config_name": "en-wa", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 87099, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 6085948, "num_examples": 104496, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 87726, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-wa-v1.0.tar.gz": {"num_bytes": 2119821, "checksum": "5b60c912eb04afe15df000ca3f3e01c33a63ced61062bb1bb79d007a78bb021d"}}, "download_size": 2119821, "post_processing_size": null, "dataset_size": 6260773, "size_in_bytes": 8380594}, "en-xh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "xh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "xh"}, "builder_name": "opus100", "config_name": "en-xh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 318660, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 50607248, "num_examples": 439671, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 315839, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-xh-v1.0.tar.gz": {"num_bytes": 20503199, "checksum": "89030efe5382cce5c1c4085651fad222e04f43487ac1082b7aeeed8b6f8e2bf4"}}, "download_size": 20503199, "post_processing_size": null, "dataset_size": 51241747, "size_in_bytes": 71744946}, "en-yi": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "yi"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "yi"}, "builder_name": "opus100", "config_name": "en-yi", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 96490, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 1275143, "num_examples": 15010, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 99826, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-yi-v1.0.tar.gz": {"num_bytes": 284031, "checksum": "d064b552fedb167dcdf9979db338e4c239cabbe238fd03755d61775c4dbb3945"}}, "download_size": 284031, "post_processing_size": null, "dataset_size": 1471459, "size_in_bytes": 1755490}, "en-yo": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "yo"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "yo"}, "builder_name": "opus100", "config_name": "en-yo", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 979769, "num_examples": 10375, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-yo-v1.0.tar.gz": {"num_bytes": 177540, "checksum": "42a74826c0c29b35255186a706c3b15b43995d9ca34b0e68aa60700be9069e06"}}, "download_size": 177540, "post_processing_size": null, "dataset_size": 979769, "size_in_bytes": 1157309}, "en-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "zh"}, "builder_name": "opus100", "config_name": "en-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 511372, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 200062983, "num_examples": 1000000, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 512364, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-zh-v1.0.tar.gz": {"num_bytes": 83265500, "checksum": "c417af397ec380032528a9de8e2fbed40e2de78e0bcc6bace0de257a01ae6635"}}, "download_size": 83265500, "post_processing_size": null, "dataset_size": 201086719, "size_in_bytes": 284352219}, "en-zu": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["en", "zu"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "en", "output": "zu"}, "builder_name": "opus100", "config_name": "en-zu", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 117518, "num_examples": 2000, "dataset_name": "opus100"}, "train": {"name": "train", "num_bytes": 2799590, "num_examples": 38616, "dataset_name": "opus100"}, "validation": {"name": "validation", "num_bytes": 120141, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-en-zu-v1.0.tar.gz": {"num_bytes": 889951, "checksum": "0b488dd9205d2801b80ae9e72a946ef1f8ecd4dbe47c54c8788c8c62b7397056"}}, "download_size": 889951, "post_processing_size": null, "dataset_size": 3037249, "size_in_bytes": 3927200}, "ar-de": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "de"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "de"}, "builder_name": "opus100", "config_name": "ar-de", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 238599, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 238599, "size_in_bytes": 2795390}, "ar-fr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "fr"}, "builder_name": "opus100", "config_name": "ar-fr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 547382, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 547382, "size_in_bytes": 3104173}, "ar-nl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "nl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "nl"}, "builder_name": "opus100", "config_name": "ar-nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 212936, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 212936, "size_in_bytes": 2769727}, "ar-ru": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "ru"}, "builder_name": "opus100", "config_name": "ar-ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 808270, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 808270, "size_in_bytes": 3365061}, "ar-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ar", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ar", "output": "zh"}, "builder_name": "opus100", "config_name": "ar-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 713412, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 713412, "size_in_bytes": 3270203}, "de-fr": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["de", "fr"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "de", "output": "fr"}, "builder_name": "opus100", "config_name": "de-fr", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 458746, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 458746, "size_in_bytes": 3015537}, "de-nl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["de", "nl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "de", "output": "nl"}, "builder_name": "opus100", "config_name": "de-nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 403886, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 403886, "size_in_bytes": 2960677}, "de-ru": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["de", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "de", "output": "ru"}, "builder_name": "opus100", "config_name": "de-ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 315779, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 315779, "size_in_bytes": 2872570}, "de-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["de", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "de", "output": "zh"}, "builder_name": "opus100", "config_name": "de-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 280397, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 280397, "size_in_bytes": 2837188}, "fr-nl": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["fr", "nl"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "fr", "output": "nl"}, "builder_name": "opus100", "config_name": "fr-nl", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 368646, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 368646, "size_in_bytes": 2925437}, "fr-ru": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["fr", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "fr", "output": "ru"}, "builder_name": "opus100", "config_name": "fr-ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 732724, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 732724, "size_in_bytes": 3289515}, "fr-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["fr", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "fr", "output": "zh"}, "builder_name": "opus100", "config_name": "fr-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 619394, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 619394, "size_in_bytes": 3176185}, "nl-ru": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["nl", "ru"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "nl", "output": "ru"}, "builder_name": "opus100", "config_name": "nl-ru", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 256067, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 256067, "size_in_bytes": 2812858}, "nl-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["nl", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "nl", "output": "zh"}, "builder_name": "opus100", "config_name": "nl-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 183641, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 183641, "size_in_bytes": 2740432}, "ru-zh": {"description": "OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.\nThe corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.\nOf the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.\n", "citation": "@misc{zhang2020improving,\n      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},\n      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},\n      year={2020},\n      eprint={2004.11867},\n      archivePrefix={arXiv},\n      primaryClass={cs.CL}\n}\n", "homepage": "http://opus.nlpl.eu/opus-100.php", "license": "", "features": {"translation": {"languages": ["ru", "zh"], "id": null, "_type": "Translation"}}, "post_processed": null, "supervised_keys": {"input": "ru", "output": "zh"}, "builder_name": "opus100", "config_name": "ru-zh", "version": {"version_str": "0.0.0", "description": null, "major": 0, "minor": 0, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 916114, "num_examples": 2000, "dataset_name": "opus100"}}, "download_checksums": {"https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz": {"num_bytes": 2556791, "checksum": "241de5bf4eb382017337f3ffb23cdf319c971585b5f679da9a420f23a5c2ed8f"}}, "download_size": 2556791, "post_processing_size": null, "dataset_size": 916114, "size_in_bytes": 3472905}}
\ No newline at end of file
diff --git a/de-en/test-00000-of-00001.parquet b/de-en/test-00000-of-00001.parquet
deleted file mode 100644
index 4905917c863c6de102e27f906217054891ad0a85..0000000000000000000000000000000000000000
--- a/de-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:05913515e9dc8c11bc03570bd00ae5b551c32b03e07901369f91372ad63a3f11
-size 253475
diff --git a/de-en/train-00000-of-00001.parquet b/de-en/train-00000-of-00001.parquet
deleted file mode 100644
index 81fbb8de62b5162aa527468da48b32f5f2d71909..0000000000000000000000000000000000000000
--- a/de-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:496a8ce873740047105f4cc9b503c94f78e8446ad02b48d24d137b54cd8f4aec
-size 116173753
diff --git a/de-en/validation-00000-of-00001.parquet b/de-en/validation-00000-of-00001.parquet
deleted file mode 100644
index 8bbbfd81df5695bdd25f4467dce45b5138d147d9..0000000000000000000000000000000000000000
--- a/de-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ae94b9183313ae36afca9c117e2be818de2700abaf57e1ad75d0111ffdcd8efe
-size 253662
diff --git a/de-fr/test-00000-of-00001.parquet b/de-fr/test-00000-of-00001.parquet
deleted file mode 100644
index 722e0dd0241bb4e2d3748abd3773d68cda108816..0000000000000000000000000000000000000000
--- a/de-fr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:efa534f446148c71fc0f9898e48071197a4dfeaf3a9b281bffa88e61227842e3
-size 311929
diff --git a/de-nl/test-00000-of-00001.parquet b/de-nl/test-00000-of-00001.parquet
deleted file mode 100644
index fad542e7eb37de6a13550d5a2a070b3f39fb17f2..0000000000000000000000000000000000000000
--- a/de-nl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:36796675dfcd8e12303b7673f9fcbdc47821e7101b7813b8998db03501ed946b
-size 281548
diff --git a/de-ru/test-00000-of-00001.parquet b/de-ru/test-00000-of-00001.parquet
deleted file mode 100644
index b597b465073b9ab0042d75dd49f2d24faac75654..0000000000000000000000000000000000000000
--- a/de-ru/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fc0f5c583e82cf3a18befdfc981eb81c8450a11f406acccb20593c14c8894a63
-size 203225
diff --git a/de-zh/test-00000-of-00001.parquet b/de-zh/test-00000-of-00001.parquet
deleted file mode 100644
index 5a1beb83aaff21e92f0b90102d042389786719e8..0000000000000000000000000000000000000000
--- a/de-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e1c97f214981d944ca1a15f4a8bee04eace89fce6f9cc8cb6c1731f1525dfd45
-size 215301
diff --git a/dz-en/train-00000-of-00001.parquet b/dz-en/train-00000-of-00001.parquet
deleted file mode 100644
index 8007a404a6e5416f161c2de7f8d242f5757a0f8c..0000000000000000000000000000000000000000
--- a/dz-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fae30597e0207036ec6212a830c1aa9590199a40b2e39b50affb1c004c17de70
-size 37361
diff --git a/el-en/test-00000-of-00001.parquet b/el-en/test-00000-of-00001.parquet
deleted file mode 100644
index 13d5009c54c236f2496acf69f653ab78c97cad35..0000000000000000000000000000000000000000
--- a/el-en/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:50023df17dbf8e0e79e18bf261ed57661a4fd5c3be8d216ea29b2904fe110b7f
-size 193700
diff --git a/el-en/train-00000-of-00001.parquet b/el-en/train-00000-of-00001.parquet
deleted file mode 100644
index 41dd1e668ea58da9123d04f5d78b4c3ba27f8a05..0000000000000000000000000000000000000000
--- a/el-en/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f2ff1e42812a294436d82431b0565750c950f17dd46e7bd384fc25744147160b
-size 83753976
diff --git a/el-en/validation-00000-of-00001.parquet b/el-en/validation-00000-of-00001.parquet
deleted file mode 100644
index a826b5e57d3c479a3f5799a6a6c83512d687a102..0000000000000000000000000000000000000000
--- a/el-en/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:546e16e1463427f9ff3567f142b391c92f89ef67ea14f115f30e1eb03bc4476d
-size 190046
diff --git a/en-eo/test-00000-of-00001.parquet b/en-eo/test-00000-of-00001.parquet
deleted file mode 100644
index 160983c58f81d76ea0394fb8335ebb69624c234a..0000000000000000000000000000000000000000
--- a/en-eo/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:72fefee2f46222f58db6dcfba15bc736f4ad542ff08986b15b84a2418226ef3e
-size 127593
diff --git a/en-eo/train-00000-of-00001.parquet b/en-eo/train-00000-of-00001.parquet
deleted file mode 100644
index b08aa3c3cd1b47465215da70f3d3379355e24621..0000000000000000000000000000000000000000
--- a/en-eo/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ef31d48c4e6a2f8ee8114a9464976372cdff2e79af6323e64fb2b6821470a040
-size 19289264
diff --git a/en-eo/validation-00000-of-00001.parquet b/en-eo/validation-00000-of-00001.parquet
deleted file mode 100644
index 0523501f401e5f84ee6df237271bd6e52cb2a314..0000000000000000000000000000000000000000
--- a/en-eo/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6b5371d84ba5103ccb1acc91bc7754ee7478841fb1db0c5e41e31b70608a0024
-size 128604
diff --git a/en-es/test-00000-of-00001.parquet b/en-es/test-00000-of-00001.parquet
deleted file mode 100644
index b63c520e95ec794df52b4e33737b82002557a4f8..0000000000000000000000000000000000000000
--- a/en-es/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7f0e843bd8050aea612350c86dd53c1eae9fc00985d9d9d1e021840491aa7290
-size 237184
diff --git a/en-es/train-00000-of-00001.parquet b/en-es/train-00000-of-00001.parquet
deleted file mode 100644
index d736a460eef66464bc9cf02d17f8d6c1a702d730..0000000000000000000000000000000000000000
--- a/en-es/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f62b9ab638576b8de12375f26b9d2f9162e1b3ac9877f6f97fe29511806ab02c
-size 99628776
diff --git a/en-es/validation-00000-of-00001.parquet b/en-es/validation-00000-of-00001.parquet
deleted file mode 100644
index 46b447774936923448e170b1f1fadbc7c5c6d14f..0000000000000000000000000000000000000000
--- a/en-es/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0630301dbcf7ef4c51cd37d6c4258728b5628d43b23bf160f9aabcb6bf79f8f0
-size 237947
diff --git a/en-et/test-00000-of-00001.parquet b/en-et/test-00000-of-00001.parquet
deleted file mode 100644
index 7111f684409cae1dbc65d4831faa93e85aa1298e..0000000000000000000000000000000000000000
--- a/en-et/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f9eee7481a6543fcd29ce65a48b522ece18c3e5f59bc9eacd85567239e32d5ae
-size 199403
diff --git a/en-et/train-00000-of-00001.parquet b/en-et/train-00000-of-00001.parquet
deleted file mode 100644
index 669ecdde8a1fd299797be4f918fc2a37e1deb18f..0000000000000000000000000000000000000000
--- a/en-et/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0c8ed77609324f55336bb9ad6a1458966fb42e7bb1ccecfacbb929265b3a4872
-size 83286767
diff --git a/en-et/validation-00000-of-00001.parquet b/en-et/validation-00000-of-00001.parquet
deleted file mode 100644
index a1ff25b105e0446b48b5e7c27aab526a0b5ce34e..0000000000000000000000000000000000000000
--- a/en-et/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:13648c3322ae272e05d5a44ea1a054604f7c05f99810f87e65cb579161029091
-size 204280
diff --git a/en-eu/test-00000-of-00001.parquet b/en-eu/test-00000-of-00001.parquet
deleted file mode 100644
index 2a223deed01dbcb4d6ffdc42e8407f727e87aa3c..0000000000000000000000000000000000000000
--- a/en-eu/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:50c54c226f2989ed35d4da5289275a7b8097197261b135bfa933a5fbf047421d
-size 208937
diff --git a/en-eu/train-00000-of-00001.parquet b/en-eu/train-00000-of-00001.parquet
deleted file mode 100644
index a8837f4b69490fedb6703be1f4a3cf8e254ece95..0000000000000000000000000000000000000000
--- a/en-eu/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:73cbe91e3405d44652cba506ff183ed5132f2d857e47d51d8641e30fbcb6cb52
-size 84388403
diff --git a/en-eu/validation-00000-of-00001.parquet b/en-eu/validation-00000-of-00001.parquet
deleted file mode 100644
index b170bcd17f1226c86031048a2cf9bc9e42d5f3a0..0000000000000000000000000000000000000000
--- a/en-eu/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cf60369c92c6b7a1186315c728eaa726c416abc71ec982f62be6710e81a6a6a5
-size 208127
diff --git a/en-fa/test-00000-of-00001.parquet b/en-fa/test-00000-of-00001.parquet
deleted file mode 100644
index fc51deebac8544c3311d8c99f93223357c372a3d..0000000000000000000000000000000000000000
--- a/en-fa/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a44d020380e314b40f292b8e68230e7454bba574e628b4190a6576d2c35f3151
-size 193412
diff --git a/en-fa/train-00000-of-00001.parquet b/en-fa/train-00000-of-00001.parquet
deleted file mode 100644
index ac4e7f6b2ac4789b16287e3151a7904662e1eb7c..0000000000000000000000000000000000000000
--- a/en-fa/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:64c4c2ee1cbcda187a5a6a55b51ec9aa96d1cc09d785ffe5d5f2fcedf3de7b1d
-size 82400160
diff --git a/en-fa/validation-00000-of-00001.parquet b/en-fa/validation-00000-of-00001.parquet
deleted file mode 100644
index 5cfa29c69ab9b1228791aac73ce4db785600fb77..0000000000000000000000000000000000000000
--- a/en-fa/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:65023eda83c8688c6183f1dd6b51ca990e1fb6e4ac70a9462d90c7aa01e2b686
-size 189676
diff --git a/en-fi/test-00000-of-00001.parquet b/en-fi/test-00000-of-00001.parquet
deleted file mode 100644
index f0b6e201e992538a2bdce9f0bcf7e7e4544c9288..0000000000000000000000000000000000000000
--- a/en-fi/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e35d33d5de5db7dcc097ddc48bb3264a307b45289b920b5b7b648b9f71be1b0b
-size 181524
diff --git a/en-fi/train-00000-of-00001.parquet b/en-fi/train-00000-of-00001.parquet
deleted file mode 100644
index 5713a1cf2e562f7f5613d07f20b0d988eb6f076b..0000000000000000000000000000000000000000
--- a/en-fi/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:da0b889ca3fd585d83f4be59c43a6b9050ac3a4d9d1ae0fe43a7cf59173866b8
-size 78957628
diff --git a/en-fi/validation-00000-of-00001.parquet b/en-fi/validation-00000-of-00001.parquet
deleted file mode 100644
index ebe2fe6f032ef48ee1e63c8c1e3ea37268dc4b96..0000000000000000000000000000000000000000
--- a/en-fi/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3ff0516368cd2e54ca00cd72111dc7c3343708691640989940cc0257b5d44a17
-size 181068
diff --git a/en-fr/test-00000-of-00001.parquet b/en-fr/test-00000-of-00001.parquet
deleted file mode 100644
index bdf41106b09200c18c479ef08aba9b5a50393e77..0000000000000000000000000000000000000000
--- a/en-fr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6e5862c14744efb89cf4c807cf0fd1a5969249935f21a1d03f3fbdbc0fb81971
-size 327139
diff --git a/en-fr/train-00000-of-00001.parquet b/en-fr/train-00000-of-00001.parquet
deleted file mode 100644
index f6a56e419a9381b8d9ff2d048ec0e8a510c2d8e7..0000000000000000000000000000000000000000
--- a/en-fr/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ddd10852ab60339e51067ec15876ae1759e4e0bedf6fbba9e8a942a099a8bcc3
-size 141590671
diff --git a/en-fr/validation-00000-of-00001.parquet b/en-fr/validation-00000-of-00001.parquet
deleted file mode 100644
index 188e12339a860fdf84288dbbd5ab5eada6025654..0000000000000000000000000000000000000000
--- a/en-fr/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6698493d99a10056d89f88c2f64af16cb8614bb9c0bf879f1cb70dd3b507bfc3
-size 334050
diff --git a/en-fy/test-00000-of-00001.parquet b/en-fy/test-00000-of-00001.parquet
deleted file mode 100644
index 324347ce2fc4f9145d5048aa4a57ab6b9b6ca789..0000000000000000000000000000000000000000
--- a/en-fy/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:933f0b9b72dbab8c308bff13203b07b9261c5cab67e85c48564d4f248e7edf91
-size 64936
diff --git a/en-fy/train-00000-of-00001.parquet b/en-fy/train-00000-of-00001.parquet
deleted file mode 100644
index 8b6681c126eb2b90bcb68b0d9f7b7d28e66a1d95..0000000000000000000000000000000000000000
--- a/en-fy/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:70dcb9a7899ec436fad0d80b43ca567acfd8741159f273c689c6da7b82664387
-size 2855742
diff --git a/en-fy/validation-00000-of-00001.parquet b/en-fy/validation-00000-of-00001.parquet
deleted file mode 100644
index 5d62b4de09e8bd27a6f8d37c140dd62bf9593cb2..0000000000000000000000000000000000000000
--- a/en-fy/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:996f44d89285a397e8eb5fcb039cba2b4b2efe0ea81cd5efebe0b4ab75bc9e73
-size 63605
diff --git a/en-ga/test-00000-of-00001.parquet b/en-ga/test-00000-of-00001.parquet
deleted file mode 100644
index e8ea42329404794ccc6987807597962635a2d452..0000000000000000000000000000000000000000
--- a/en-ga/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:12d287c79745c5cdc43e7bc2d635735368b16d2eea1a0d5c7b200bc37f39d788
-size 299428
diff --git a/en-ga/train-00000-of-00001.parquet b/en-ga/train-00000-of-00001.parquet
deleted file mode 100644
index d96aaf75c7cbe82eb368b973b82f0cf27d1bbecc..0000000000000000000000000000000000000000
--- a/en-ga/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:37fb92e45685885bfff8f55cd6c1c3d984806068081a3d04eee8c39ff854ba52
-size 27339840
diff --git a/en-ga/validation-00000-of-00001.parquet b/en-ga/validation-00000-of-00001.parquet
deleted file mode 100644
index 55e565f6aba0707e18a07a83cd36f483d6e10ecf..0000000000000000000000000000000000000000
--- a/en-ga/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:58fe26cafefba27de9abdf12a7214e3cb25f1062e3bb9a42d202058ccc5ace37
-size 298180
diff --git a/en-gd/test-00000-of-00001.parquet b/en-gd/test-00000-of-00001.parquet
deleted file mode 100644
index 67e73fc43ae3dd0206b54c4f1345a19a6140c73a..0000000000000000000000000000000000000000
--- a/en-gd/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:dcf27275b79d8732274460934a1c780d7ebff030175fc615c50e8e51779d042f
-size 121521
diff --git a/en-gd/train-00000-of-00001.parquet b/en-gd/train-00000-of-00001.parquet
deleted file mode 100644
index 1748b7b1c83fac6b0ca50bd4270834d019aa2c1f..0000000000000000000000000000000000000000
--- a/en-gd/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2011ff662d27109f5563db82926237b92a356acc49d04dcb682c5ecef538cf5a
-size 888333
diff --git a/en-gd/validation-00000-of-00001.parquet b/en-gd/validation-00000-of-00001.parquet
deleted file mode 100644
index 3ce734bd7d851845ad36000a031b97f064795309..0000000000000000000000000000000000000000
--- a/en-gd/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ead688063a130174077e105800c98bc7086ec348b70adccc1f34ee15f09fdd31
-size 114652
diff --git a/en-gl/test-00000-of-00001.parquet b/en-gl/test-00000-of-00001.parquet
deleted file mode 100644
index 84b618b57981e47c59dc127f7223221ce601e1a0..0000000000000000000000000000000000000000
--- a/en-gl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4f4b7eb9f0d5dcdbc879423fe759c363e5cf31a5cf84409240fabeb039df3f66
-size 145791
diff --git a/en-gl/train-00000-of-00001.parquet b/en-gl/train-00000-of-00001.parquet
deleted file mode 100644
index 107c494a0cf5a13f80a0d6ca5dec611690ec339f..0000000000000000000000000000000000000000
--- a/en-gl/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ec7b246b32c2a9cd90dad4fb77edcc4027a88dfacab00dcc4241b99b591ef354
-size 33789002
diff --git a/en-gl/validation-00000-of-00001.parquet b/en-gl/validation-00000-of-00001.parquet
deleted file mode 100644
index b8eb5f12a4ef5c2e261f21ed232200ae843870dc..0000000000000000000000000000000000000000
--- a/en-gl/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:29cc088b8ee77db5fe333e99db991f176d693617e512ac322b791ce24c28ac98
-size 149235
diff --git a/en-gu/test-00000-of-00001.parquet b/en-gu/test-00000-of-00001.parquet
deleted file mode 100644
index 6dbc5c26b7a409a369ed2f29339194598a581c7c..0000000000000000000000000000000000000000
--- a/en-gu/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7f0d5a42f247a13c38be2ba46679279ca760590e554abc171a50b93848b3ebc9
-size 103986
diff --git a/en-gu/train-00000-of-00001.parquet b/en-gu/train-00000-of-00001.parquet
deleted file mode 100644
index 836818cae35297565ee7bb6b3b7aba635ae15ef1..0000000000000000000000000000000000000000
--- a/en-gu/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:716669a675592e6fc8ccf672f472ece03b92e073e5a27de69dd38ea6cb90d5cf
-size 19024532
diff --git a/en-gu/validation-00000-of-00001.parquet b/en-gu/validation-00000-of-00001.parquet
deleted file mode 100644
index 4a25c9c79e64dceb1dae25a800cea51e8bd462a5..0000000000000000000000000000000000000000
--- a/en-gu/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a17404f2f5f24333956aea6d8ba6192bf744eed08fba6492fc668161b791fba0
-size 107261
diff --git a/en-ha/test-00000-of-00001.parquet b/en-ha/test-00000-of-00001.parquet
deleted file mode 100644
index fd24f877d5e41f1bccc58fe3b9f1d611bfcbf9ab..0000000000000000000000000000000000000000
--- a/en-ha/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5a2de46d58abb9eb09b2e6b173d00799aea4eed009bbd647eb3941aad94333ce
-size 244173
diff --git a/en-ha/train-00000-of-00001.parquet b/en-ha/train-00000-of-00001.parquet
deleted file mode 100644
index 26c134b83cf60f548cc23de087505e5181548ea9..0000000000000000000000000000000000000000
--- a/en-ha/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2ff971cb32a199418dbeef9f769589630268dad10cf0240a4d8d4eb75488edc7
-size 12195854
diff --git a/en-ha/validation-00000-of-00001.parquet b/en-ha/validation-00000-of-00001.parquet
deleted file mode 100644
index 7a2a47b1c32f087d83febc214890ee61a2ba8307..0000000000000000000000000000000000000000
--- a/en-ha/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3c1fc5e399e1fd7f8acd3d78d0273e64804a965d6b674d9a6639d18e7b64cd32
-size 246160
diff --git a/en-he/test-00000-of-00001.parquet b/en-he/test-00000-of-00001.parquet
deleted file mode 100644
index 377edd53b90ba214625fa2080bffcb3c202efb30..0000000000000000000000000000000000000000
--- a/en-he/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:570a1fbd7bc1b7cde0af1abf0fe83cfc1fe9b8290c9869efebb5f0ceee4e534a
-size 139237
diff --git a/en-he/train-00000-of-00001.parquet b/en-he/train-00000-of-00001.parquet
deleted file mode 100644
index 655762ddf51b3c5d953d41b964da0b5e16b503eb..0000000000000000000000000000000000000000
--- a/en-he/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:28994d37e69796aaddaed9d71060a448592d8c8eb9906f7f39cd1661ab1afa4a
-size 60865553
diff --git a/en-he/validation-00000-of-00001.parquet b/en-he/validation-00000-of-00001.parquet
deleted file mode 100644
index 2bbf7980b449cc25811baa333f48c448425962fb..0000000000000000000000000000000000000000
--- a/en-he/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:66c2f904d55dd3071f56aea4d83c69cdd73b871eeadb263c36e14898b368eea8
-size 139968
diff --git a/en-hi/test-00000-of-00001.parquet b/en-hi/test-00000-of-00001.parquet
deleted file mode 100644
index 3a79a2a1c9954b4425a19c126e0737282e5f2675..0000000000000000000000000000000000000000
--- a/en-hi/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9d0caf331e4e9dd6fee478622a437bf9ec833c6af6193cf665d48ac1d746caaa
-size 259276
diff --git a/en-hi/train-00000-of-00001.parquet b/en-hi/train-00000-of-00001.parquet
deleted file mode 100644
index b6528a043d0775f0d5c86d5292acaf41078e214d..0000000000000000000000000000000000000000
--- a/en-hi/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5c3a47dd694f009f4cb63d467bc38e7ce0b6c75de847b69e866d9d998178f916
-size 65219235
diff --git a/en-hi/validation-00000-of-00001.parquet b/en-hi/validation-00000-of-00001.parquet
deleted file mode 100644
index 3795a0d026e7f9762e6d213cd288166c0959358d..0000000000000000000000000000000000000000
--- a/en-hi/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b94a11774c4a691f043400cedfff637fc9b39d26f02d5d3d30611840c11dc931
-size 247375
diff --git a/en-hr/test-00000-of-00001.parquet b/en-hr/test-00000-of-00001.parquet
deleted file mode 100644
index 53500d4a6e66d5579c8cfa8237c16e9294591bed..0000000000000000000000000000000000000000
--- a/en-hr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:54deb07dadf324453b0d30913fc6dc9d2a41671c990f4bc79544a90f332d3586
-size 141338
diff --git a/en-hr/train-00000-of-00001.parquet b/en-hr/train-00000-of-00001.parquet
deleted file mode 100644
index 090fa0f68ab847e2addde1bb941c3b31868b50a0..0000000000000000000000000000000000000000
--- a/en-hr/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c8cf607c9f38aa78f624ffb3bf0f7b4b728142098beed20f9ce2fabfbaa4bb12
-size 59186522
diff --git a/en-hr/validation-00000-of-00001.parquet b/en-hr/validation-00000-of-00001.parquet
deleted file mode 100644
index 077bcafafc21e7694de7f1cdd2a0fb9b1605ba1d..0000000000000000000000000000000000000000
--- a/en-hr/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7044cd7fd485af20d5ce8c6a68a673ff61be79620782049f0d63f694fb267be8
-size 141032
diff --git a/en-hu/test-00000-of-00001.parquet b/en-hu/test-00000-of-00001.parquet
deleted file mode 100644
index 8c00eb4fb922578e2c7b15577433b3e7d8352307..0000000000000000000000000000000000000000
--- a/en-hu/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:699dc0c1e88dc1185b07771837c7b1a6a0369b8be8d4120849164de2a3ca7a4e
-size 156704
diff --git a/en-hu/train-00000-of-00001.parquet b/en-hu/train-00000-of-00001.parquet
deleted file mode 100644
index 5d5e57e60689da74cfc5031c35118a80a4359a38..0000000000000000000000000000000000000000
--- a/en-hu/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c6bb3733f9f195940dec1c92f524ba07bb9a28f9275e31eab3571a056c6aacbf
-size 67654965
diff --git a/en-hu/validation-00000-of-00001.parquet b/en-hu/validation-00000-of-00001.parquet
deleted file mode 100644
index 121352b650ebdef8e13e9c5a265edb91bef370bb..0000000000000000000000000000000000000000
--- a/en-hu/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:79fe377fcd06d3ea601b3dfbe3fdee83ac8c28ca069881cdd028d79fb0b16dce
-size 159447
diff --git a/en-hy/train-00000-of-00001.parquet b/en-hy/train-00000-of-00001.parquet
deleted file mode 100644
index da91925490ed087d2ef94a8280fd63fbbf18a355..0000000000000000000000000000000000000000
--- a/en-hy/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0db2dcda454d2b7f6950dc56590cc6bdfead931486fa7adf031982b0639b3c4a
-size 422847
diff --git a/en-id/test-00000-of-00001.parquet b/en-id/test-00000-of-00001.parquet
deleted file mode 100644
index 26529b0c2dac954d5fd2271b576da8c6ddbd0c18..0000000000000000000000000000000000000000
--- a/en-id/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ba22dcb562f57b345caae8c1801ad06ca9836f9e33b8e651277da2906aa3ccb2
-size 129317
diff --git a/en-id/train-00000-of-00001.parquet b/en-id/train-00000-of-00001.parquet
deleted file mode 100644
index 779ab0d01ff9fbdbedcdd804b14c77410a761431..0000000000000000000000000000000000000000
--- a/en-id/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fa02ba8bc9e8952a134a4705b4a7af8f1301907a5447b00f2e11abf0ba842900
-size 57432415
diff --git a/en-id/validation-00000-of-00001.parquet b/en-id/validation-00000-of-00001.parquet
deleted file mode 100644
index 8775945cae66ffe83743d6278943625fd18a308d..0000000000000000000000000000000000000000
--- a/en-id/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ff5448f86e52784f465af03a5b3de4e54eeb86e2938cf20685e431648083d3a0
-size 131946
diff --git a/en-ig/test-00000-of-00001.parquet b/en-ig/test-00000-of-00001.parquet
deleted file mode 100644
index e4e423931b2cb102c0dd0e19d55c0362b933dbdb..0000000000000000000000000000000000000000
--- a/en-ig/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f3a32940b625bb0b5aeee66311b3fadf71fe7745639692a85a169cf54862c637
-size 44351
diff --git a/en-ig/train-00000-of-00001.parquet b/en-ig/train-00000-of-00001.parquet
deleted file mode 100644
index 37bd703160b400c4e39a15076d4d66d49250d396..0000000000000000000000000000000000000000
--- a/en-ig/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:804e6b0f243d02ed7a00367450e3a67e2198c894d20c604c11b10f44417736c6
-size 769769
diff --git a/en-ig/validation-00000-of-00001.parquet b/en-ig/validation-00000-of-00001.parquet
deleted file mode 100644
index 7747c3def627a545c1df2ab76023e41437713864..0000000000000000000000000000000000000000
--- a/en-ig/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2250c55cb1ee037c535d751e25c00ee4af689473dc38035f1e177ac9a1537bd0
-size 45320
diff --git a/en-is/test-00000-of-00001.parquet b/en-is/test-00000-of-00001.parquet
deleted file mode 100644
index 5aed43efbc7d645f59df2c19cec8c2ae0f29a155..0000000000000000000000000000000000000000
--- a/en-is/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e5c18f4f693416294fc09a793d993cc7dc9ec8e85c62269aef2ff3d5a9836558
-size 127566
diff --git a/en-is/train-00000-of-00001.parquet b/en-is/train-00000-of-00001.parquet
deleted file mode 100644
index 60c123a2c6aab98322b18eb726bc443cb2b9360b..0000000000000000000000000000000000000000
--- a/en-is/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:14e6e9ebc7886497882ad078f9d9d2b9e2c57f210f24ef02f50b4b9455358382
-size 55986734
diff --git a/en-is/validation-00000-of-00001.parquet b/en-is/validation-00000-of-00001.parquet
deleted file mode 100644
index 559433f2f10f3fb64f09503c176a5e6d2f4a236a..0000000000000000000000000000000000000000
--- a/en-is/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c57d71573a928993dd0d0eb30e91c3371337384ebbb8f33afe1ce544bc43f5ad
-size 127849
diff --git a/en-it/test-00000-of-00001.parquet b/en-it/test-00000-of-00001.parquet
deleted file mode 100644
index 6b3387a652b0688354c6317fd0a86f82a1b48d01..0000000000000000000000000000000000000000
--- a/en-it/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:21f5e5e385267b51fac6180f0706edacb4f2e712422872a07405694e394a203c
-size 223489
diff --git a/en-it/train-00000-of-00001.parquet b/en-it/train-00000-of-00001.parquet
deleted file mode 100644
index 2e9907c4595adcf23007166fd04f62685374516f..0000000000000000000000000000000000000000
--- a/en-it/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:621615f937597dc7a37beb19fb7cd12ac0503731a513c96cab0d69ee791b37e9
-size 91690686
diff --git a/en-it/validation-00000-of-00001.parquet b/en-it/validation-00000-of-00001.parquet
deleted file mode 100644
index a474b20432fcbdd76e12e428a1f338131e382892..0000000000000000000000000000000000000000
--- a/en-it/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9c3578974ce9a9f700a8a5587e4871beb568ec71dc4befeef5f23a5be2e9bd9d
-size 219722
diff --git a/en-ja/test-00000-of-00001.parquet b/en-ja/test-00000-of-00001.parquet
deleted file mode 100644
index 556615cb2d9ca4115c9cbaad135d9aaa7f2792e2..0000000000000000000000000000000000000000
--- a/en-ja/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:74df993d8c835747279f066a5df36e691149de7656498656820b6d2f7b4219ce
-size 139878
diff --git a/en-ja/train-00000-of-00001.parquet b/en-ja/train-00000-of-00001.parquet
deleted file mode 100644
index e6b79fdc1b451f7c05ffacc79c8848b1f793457e..0000000000000000000000000000000000000000
--- a/en-ja/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:26c3812774074c1197cc2f5b320789e941efd80f94ead09814f483d11bf334ca
-size 64537336
diff --git a/en-ja/validation-00000-of-00001.parquet b/en-ja/validation-00000-of-00001.parquet
deleted file mode 100644
index 20066e8972c9fb1090a11569b8f63209225ca7a0..0000000000000000000000000000000000000000
--- a/en-ja/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e0c898f4785a1f1944ddeefe59315d4a452dfe3f3321096b6c256f0735e57ce3
-size 139894
diff --git a/en-ka/test-00000-of-00001.parquet b/en-ka/test-00000-of-00001.parquet
deleted file mode 100644
index ac994e1c6c94457db8956c8f9723146eda79d01d..0000000000000000000000000000000000000000
--- a/en-ka/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:89ccd9cc655c1641cc0a070f307cf3fc1694dd16d0d2db1ea3aed3225819b07a
-size 143273
diff --git a/en-ka/train-00000-of-00001.parquet b/en-ka/train-00000-of-00001.parquet
deleted file mode 100644
index 2d0a553e531c57d9d4a47d07a5be41bcc9d3e406..0000000000000000000000000000000000000000
--- a/en-ka/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4e80ed7f7d7ec2f6df501f0c9a2bbb01c8737ad2509cf8b59201460cbd6cdc4d
-size 24105501
diff --git a/en-ka/validation-00000-of-00001.parquet b/en-ka/validation-00000-of-00001.parquet
deleted file mode 100644
index 3b408fc0cb740b8f4e7e1a00175381cd67945776..0000000000000000000000000000000000000000
--- a/en-ka/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:eb5845cee7fc3b2aebf21380dcab3a6ddb200f7cce2aed967c88008211944f5c
-size 145859
diff --git a/en-kk/test-00000-of-00001.parquet b/en-kk/test-00000-of-00001.parquet
deleted file mode 100644
index 073039e4ff75c5f4f22632279f5de84d540cce8a..0000000000000000000000000000000000000000
--- a/en-kk/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7b1bf0ea4a3d5d60d6b3b70c067818eba3683af265673e5c0d8581896725df82
-size 84062
diff --git a/en-kk/train-00000-of-00001.parquet b/en-kk/train-00000-of-00001.parquet
deleted file mode 100644
index 39931d08ec3409cdbc232ccfff1d9081592ad3a0..0000000000000000000000000000000000000000
--- a/en-kk/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:423e5a2b0bf21870e0725ce28b8a1146a78781b4f73a353b1dd57e747a088635
-size 4641227
diff --git a/en-kk/validation-00000-of-00001.parquet b/en-kk/validation-00000-of-00001.parquet
deleted file mode 100644
index e4e9cebea4868522c87a408c05107917b04fe3b8..0000000000000000000000000000000000000000
--- a/en-kk/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:be75d772a7ec6c6ead860f5636f15c4cab6edf4e390e1cc63ee8a132fdff9235
-size 83071
diff --git a/en-km/test-00000-of-00001.parquet b/en-km/test-00000-of-00001.parquet
deleted file mode 100644
index 6c415842400647628cd4937ed8c8c85eb7214d92..0000000000000000000000000000000000000000
--- a/en-km/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d14657e5f16ce7f770490e9a32d22472cc9734bdd398e559261f527d8519671c
-size 140158
diff --git a/en-km/train-00000-of-00001.parquet b/en-km/train-00000-of-00001.parquet
deleted file mode 100644
index 9cce03dfc3c05e6097be24330c65e047a1296886..0000000000000000000000000000000000000000
--- a/en-km/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8ee5c5131916b25d4817ed32543f119a45b43f674a955ff0bef560eabaa3655a
-size 9741316
diff --git a/en-km/validation-00000-of-00001.parquet b/en-km/validation-00000-of-00001.parquet
deleted file mode 100644
index 96a58cfd6e1cf09353e5156e805838ac1c665beb..0000000000000000000000000000000000000000
--- a/en-km/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:60a83f282016615da5cac6ea2dee42cad7e5f7c816bf4b2d0cd1aa768342aff0
-size 141445
diff --git a/en-kn/test-00000-of-00001.parquet b/en-kn/test-00000-of-00001.parquet
deleted file mode 100644
index 54e8b921aab895fbbf126cf00d571c3286633988..0000000000000000000000000000000000000000
--- a/en-kn/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6ae0581fa842ae753619e02c8ecb36b799816bd0981f1164cb2551622808102f
-size 38692
diff --git a/en-kn/train-00000-of-00001.parquet b/en-kn/train-00000-of-00001.parquet
deleted file mode 100644
index 57299120b22eca09cc3dbb77e6bcd81a57849019..0000000000000000000000000000000000000000
--- a/en-kn/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:994ab9e6502203e8ee23b3b6f8bb848f44342539487fca7677b1c945da1afd9e
-size 985231
diff --git a/en-kn/validation-00000-of-00001.parquet b/en-kn/validation-00000-of-00001.parquet
deleted file mode 100644
index 5347022f9986216cba444743ef06092868b8539a..0000000000000000000000000000000000000000
--- a/en-kn/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ead53526115c4d594c33fa30fbbbead6571cedeb49599b7bd48c4aafae0d3ab6
-size 38631
diff --git a/en-ko/test-00000-of-00001.parquet b/en-ko/test-00000-of-00001.parquet
deleted file mode 100644
index 7a2c17fcc2d70441346d8ffffd6400104a80fe2e..0000000000000000000000000000000000000000
--- a/en-ko/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:95cda292579fc495e19176e6b40d16030cf6cfa00ae76cb11c9e537a8f3f0ad7
-size 143342
diff --git a/en-ko/train-00000-of-00001.parquet b/en-ko/train-00000-of-00001.parquet
deleted file mode 100644
index b6ac41f62532fa38423a41b706f41a32398606b6..0000000000000000000000000000000000000000
--- a/en-ko/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:80761982f72154d69a2ae53c003c853dc41dc4a92147a4ba903f13041b37a681
-size 70096120
diff --git a/en-ko/validation-00000-of-00001.parquet b/en-ko/validation-00000-of-00001.parquet
deleted file mode 100644
index edaeb4142e0b24cf2e04e2be3ff61123f7e804f5..0000000000000000000000000000000000000000
--- a/en-ko/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fccaaa33f08fc80dc332e983b7e7abe59e4360aff32e27df03071993086a088a
-size 143809
diff --git a/en-ku/test-00000-of-00001.parquet b/en-ku/test-00000-of-00001.parquet
deleted file mode 100644
index 90b483fcab82395415b52eb8e2d6dce2d0593308..0000000000000000000000000000000000000000
--- a/en-ku/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fc4c027aba8985b0c3b2ebfa204442d1b380bc4367f75ca0f8806e104f23bcf3
-size 141060
diff --git a/en-ku/train-00000-of-00001.parquet b/en-ku/train-00000-of-00001.parquet
deleted file mode 100644
index ce619c40ae7500440fca6424a0b6c86f40656b2b..0000000000000000000000000000000000000000
--- a/en-ku/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:070581197437dede5b07d15eecfe30fb10506dafbe89dc8861e700a6598f2bae
-size 25081500
diff --git a/en-ku/validation-00000-of-00001.parquet b/en-ku/validation-00000-of-00001.parquet
deleted file mode 100644
index 90e206252e8a2497b114caf6532c2820a459d209..0000000000000000000000000000000000000000
--- a/en-ku/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a49ad8a92f6789e9e21c28c8a0eb2827f411bca1348a4f0fa113d206b5eb2369
-size 135829
diff --git a/en-ky/test-00000-of-00001.parquet b/en-ky/test-00000-of-00001.parquet
deleted file mode 100644
index a00899b01310a60558414d9e2c5857dee16802fc..0000000000000000000000000000000000000000
--- a/en-ky/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7b7753227a501b8e607a067edee9592440b6bcc11e116bf02fa87982d8879008
-size 71691
diff --git a/en-ky/train-00000-of-00001.parquet b/en-ky/train-00000-of-00001.parquet
deleted file mode 100644
index db4cd7c02d07af00334eb84e11063214dc1dc5e1..0000000000000000000000000000000000000000
--- a/en-ky/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:896b6eb31962e4258b6313dde19b72bc9602282709266a9e636e68e6828a63fb
-size 1196120
diff --git a/en-ky/validation-00000-of-00001.parquet b/en-ky/validation-00000-of-00001.parquet
deleted file mode 100644
index 40eef47ed7d6602765165acecefa0118aac95124..0000000000000000000000000000000000000000
--- a/en-ky/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cdb50d65e56aab6c0579c3268b1bfa57f6388ce67dea590a3aaac7372d31a1a7
-size 70875
diff --git a/en-li/test-00000-of-00001.parquet b/en-li/test-00000-of-00001.parquet
deleted file mode 100644
index a131d104aabe9b5e7131514616f597b4758461ed..0000000000000000000000000000000000000000
--- a/en-li/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5136797d3b1be1b9db7cc1e2f58d96732dce193e2be85f0ebdc805b650e41008
-size 38953
diff --git a/en-li/train-00000-of-00001.parquet b/en-li/train-00000-of-00001.parquet
deleted file mode 100644
index 076b92340baea00dbf54171fe62e4ffb1f6fb726..0000000000000000000000000000000000000000
--- a/en-li/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5fcb9b48b9f842b72538465548b57831ac1639fd74f847bd5c8d988404e37e47
-size 961922
diff --git a/en-li/validation-00000-of-00001.parquet b/en-li/validation-00000-of-00001.parquet
deleted file mode 100644
index 670fe175efe075b4247b912d6aa2be5d4c553a07..0000000000000000000000000000000000000000
--- a/en-li/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7be764151357f1d5ef1089de49524ede0ef6b1421cd53520980fc78699cdcdcf
-size 39885
diff --git a/en-lt/test-00000-of-00001.parquet b/en-lt/test-00000-of-00001.parquet
deleted file mode 100644
index 730d302bf503e8620b348716ae5a5911849e376c..0000000000000000000000000000000000000000
--- a/en-lt/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d6cc56432aed4a7157f4d5045a3e6fcf5111d828c3ca1cbfe8467a769f1f3e55
-size 329009
diff --git a/en-lt/train-00000-of-00001.parquet b/en-lt/train-00000-of-00001.parquet
deleted file mode 100644
index fec4ec3274f960bc562d6be94305356d9cb5815c..0000000000000000000000000000000000000000
--- a/en-lt/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:115db8c28b056a152cf15e1d794b1cc6a3becb4be0402b714f61d66ae275cd42
-size 123796359
diff --git a/en-lt/validation-00000-of-00001.parquet b/en-lt/validation-00000-of-00001.parquet
deleted file mode 100644
index a832a27dd2806a60212a6d7e0d9e453642161a70..0000000000000000000000000000000000000000
--- a/en-lt/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3bd586f792cad310801a1e477b1e896c8b92313a0e31fd5d93895b20436781f4
-size 318685
diff --git a/en-lv/test-00000-of-00001.parquet b/en-lv/test-00000-of-00001.parquet
deleted file mode 100644
index 1024631b407c720287c43fe8efc09b8f26a13fc8..0000000000000000000000000000000000000000
--- a/en-lv/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:05118136b5f5018217b1b555cc94997ec39bad3e8d3d8fabb708bed8b9f638bc
-size 359565
diff --git a/en-lv/train-00000-of-00001.parquet b/en-lv/train-00000-of-00001.parquet
deleted file mode 100644
index 8dfcb5583fbb1aed4416e83638500a928f8097f9..0000000000000000000000000000000000000000
--- a/en-lv/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c6344a145be6783564a25f13c8ce83669f9c037151392c7fe4830571062e0f0d
-size 139829929
diff --git a/en-lv/validation-00000-of-00001.parquet b/en-lv/validation-00000-of-00001.parquet
deleted file mode 100644
index c875a583938151a71e67262be3f2a8e468af583a..0000000000000000000000000000000000000000
--- a/en-lv/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:80ec5933d1fe9bff118999efc640b2f4f68e91168aaf9a539ffc77320c1f01e3
-size 349033
diff --git a/en-mg/test-00000-of-00001.parquet b/en-mg/test-00000-of-00001.parquet
deleted file mode 100644
index 59174ada9a6de91dac2e6c24bb6414c61ce472b8..0000000000000000000000000000000000000000
--- a/en-mg/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e40a9136e19c4823dced25b9c6546f9c2db86f9c1ed486e5f2a5f2ba830fd260
-size 357235
diff --git a/en-mg/train-00000-of-00001.parquet b/en-mg/train-00000-of-00001.parquet
deleted file mode 100644
index 140885c04f72b1cfafefe96f9415acfbf0eaa24d..0000000000000000000000000000000000000000
--- a/en-mg/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e338ade40618ead531a9995b932a53ee6366b39691489c14fa5c1ca3fe086899
-size 90395977
diff --git a/en-mg/validation-00000-of-00001.parquet b/en-mg/validation-00000-of-00001.parquet
deleted file mode 100644
index 40576cda28cc87087f65371249a5305fe6c1a6c2..0000000000000000000000000000000000000000
--- a/en-mg/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7c8e71a1d506046566941388c157d37758604cf401686bb084becf1162f9317e
-size 348953
diff --git a/en-mk/test-00000-of-00001.parquet b/en-mk/test-00000-of-00001.parquet
deleted file mode 100644
index 194d6a3c87f647cd000303bd16665a125493595a..0000000000000000000000000000000000000000
--- a/en-mk/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:485ef4637e96eac21db77e7183cada4e06f76a0fe77c0d7b137dc860dd8a5ab5
-size 196212
diff --git a/en-mk/train-00000-of-00001.parquet b/en-mk/train-00000-of-00001.parquet
deleted file mode 100644
index 4439caff44326b009c63f7b7f53c589f5afd191e..0000000000000000000000000000000000000000
--- a/en-mk/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:64a47580b4cb31d171314b216dabc05667d5b4c54e3ca39fd250c0e57af7df18
-size 76418660
diff --git a/en-mk/validation-00000-of-00001.parquet b/en-mk/validation-00000-of-00001.parquet
deleted file mode 100644
index 5f5acb4e8cbc548c83ccf068f444fcde5f34208c..0000000000000000000000000000000000000000
--- a/en-mk/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f6c7656cc31a2e45e6346bdc82c7c1e9df3d74f3d9d781331277bf3d590784ca
-size 195939
diff --git a/en-ml/test-00000-of-00001.parquet b/en-ml/test-00000-of-00001.parquet
deleted file mode 100644
index 68992b799dd549b4706e93209cd59665a2385d23..0000000000000000000000000000000000000000
--- a/en-ml/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9492e03d0380465c828c3bd5c22755898fe0247e63c7a13aaad6e39e353fa91c
-size 176318
diff --git a/en-ml/train-00000-of-00001.parquet b/en-ml/train-00000-of-00001.parquet
deleted file mode 100644
index f83f9cd881414de93475f2ed63027c7da601f0a8..0000000000000000000000000000000000000000
--- a/en-ml/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3eea95982ec05c6cf12bcad7e59482a8849de27f0850e12067e4ab7be0aabbbe
-size 95148478
diff --git a/en-ml/validation-00000-of-00001.parquet b/en-ml/validation-00000-of-00001.parquet
deleted file mode 100644
index b37e9ca94741726016494f3b4da95727c2bbb61a..0000000000000000000000000000000000000000
--- a/en-ml/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9f37ba5d82d341433bf6b4adfd3398acb57bc2c561b95e826c60ccd392931307
-size 172686
diff --git a/en-mn/train-00000-of-00001.parquet b/en-mn/train-00000-of-00001.parquet
deleted file mode 100644
index 310f2ffeaf6a4ee13b73a1660a30611f6642efa4..0000000000000000000000000000000000000000
--- a/en-mn/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:89786f8b052e9c47733c53fa49cfa5ab2f5fcefc9ac7b9ab261e3812121b3f0a
-size 85037
diff --git a/en-mr/test-00000-of-00001.parquet b/en-mr/test-00000-of-00001.parquet
deleted file mode 100644
index dd6a5db3cf4dee8591c4ffa6dc7c0a09e3dc0112..0000000000000000000000000000000000000000
--- a/en-mr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8040c8449cd8ed5cebbeb9f48b89c54041f7762ee86e2e35232c3ddb0507609a
-size 128192
diff --git a/en-mr/train-00000-of-00001.parquet b/en-mr/train-00000-of-00001.parquet
deleted file mode 100644
index 166696ce7d0ee64a5e91e837655acd6efd57ee7f..0000000000000000000000000000000000000000
--- a/en-mr/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:924ff1ee87f18d08d8dee457af1742b60e6b777de2bb396e83fe93ab4fe70ad0
-size 1582935
diff --git a/en-mr/validation-00000-of-00001.parquet b/en-mr/validation-00000-of-00001.parquet
deleted file mode 100644
index 0fead9608bd392d7eaade188666ae90b398e3184..0000000000000000000000000000000000000000
--- a/en-mr/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4db2cb2978663f59275d152860b562d86d8eda71066e81db685514799a1b5470
-size 127491
diff --git a/en-ms/test-00000-of-00001.parquet b/en-ms/test-00000-of-00001.parquet
deleted file mode 100644
index d08375dcb96f4ffa913185d3676ac4c9842aaa52..0000000000000000000000000000000000000000
--- a/en-ms/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b0913b3413d1cebcedfd3907bc31bd3f79233c80fa150ffa3ae5ddccbce27ec3
-size 131945
diff --git a/en-ms/train-00000-of-00001.parquet b/en-ms/train-00000-of-00001.parquet
deleted file mode 100644
index 5ff5e1ac51d6c4c411ede9b3c258a0cc91c4b459..0000000000000000000000000000000000000000
--- a/en-ms/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4bc613f9d9dd129e49b8e42b87a72404de3a3c7e1e9b1d66614e91e721108069
-size 57148882
diff --git a/en-ms/validation-00000-of-00001.parquet b/en-ms/validation-00000-of-00001.parquet
deleted file mode 100644
index 0cc0989e82843322a714224dc0cb84f1b55769fa..0000000000000000000000000000000000000000
--- a/en-ms/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:27853555a51d4b2603dc2ebc70a2be4c9d24ed3bc3f5331cda68471700d60b09
-size 132009
diff --git a/en-mt/test-00000-of-00001.parquet b/en-mt/test-00000-of-00001.parquet
deleted file mode 100644
index 2ba56f539790aec1e6c098e676316c50af88668c..0000000000000000000000000000000000000000
--- a/en-mt/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9c7d129e0c53d5055894fd25bfe3a3a2402f9fbe0485db421105fa58608755ed
-size 368212
diff --git a/en-mt/train-00000-of-00001.parquet b/en-mt/train-00000-of-00001.parquet
deleted file mode 100644
index 68b1298afee9c29525daef0fed543dad6251a694..0000000000000000000000000000000000000000
--- a/en-mt/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:23510b8ffb3f5ee6e3854e1ffdf3e38910d2b03285b89eef435eed3841a63769
-size 147083163
diff --git a/en-mt/validation-00000-of-00001.parquet b/en-mt/validation-00000-of-00001.parquet
deleted file mode 100644
index 554555821575210e43edd4e39d7fe4739440ac29..0000000000000000000000000000000000000000
--- a/en-mt/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b7c5114c681d1abe2ed26d67c6073a4576badae0f5fed8b7b5a1c6f218337d8f
-size 385262
diff --git a/en-my/test-00000-of-00001.parquet b/en-my/test-00000-of-00001.parquet
deleted file mode 100644
index 9ae84c6567ff0dde43851a391f7ba4ae298dbb3d..0000000000000000000000000000000000000000
--- a/en-my/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:731163a1db72017fc81368b0ff6e3f5e79919e21c90945f911db5d48248cd1e3
-size 139903
diff --git a/en-my/train-00000-of-00001.parquet b/en-my/train-00000-of-00001.parquet
deleted file mode 100644
index faecb97517aff2c83c4f6483dfcadb00fd279b06..0000000000000000000000000000000000000000
--- a/en-my/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5325dd25f5473934e498b1ddaf01f2166732334be8fe73de815d113fe3712bf6
-size 1675487
diff --git a/en-my/validation-00000-of-00001.parquet b/en-my/validation-00000-of-00001.parquet
deleted file mode 100644
index 09a3c57aa1efb1641fbcd5a51ec9bfc950c3d0c1..0000000000000000000000000000000000000000
--- a/en-my/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2348dda10ce46bd9bf224d8db07432c6ad02f52a8fe5ac0e3fbb8b6227195b34
-size 137183
diff --git a/en-nb/test-00000-of-00001.parquet b/en-nb/test-00000-of-00001.parquet
deleted file mode 100644
index ec406e8e8db990d9663715e3cbcfe16f5a9df9de..0000000000000000000000000000000000000000
--- a/en-nb/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9da91f0844cab38db49ff734d01615e293f81340ad70145b19a5a03a116df6fa
-size 223209
diff --git a/en-nb/train-00000-of-00001.parquet b/en-nb/train-00000-of-00001.parquet
deleted file mode 100644
index 0d6040160c33adb7795155e12293a8959be8625a..0000000000000000000000000000000000000000
--- a/en-nb/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4c56e8d0ff904c36f1fd696086ba14744e5e01962e80ccea2a636d471c92dbf4
-size 10189518
diff --git a/en-nb/validation-00000-of-00001.parquet b/en-nb/validation-00000-of-00001.parquet
deleted file mode 100644
index 5d6eb19a3cf2f16c2c3eec46d311dc82a8592ec5..0000000000000000000000000000000000000000
--- a/en-nb/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:514a1fc24fb1b8ae682336426ae624bd76a0aafb0f22905656417b83624596e0
-size 218042
diff --git a/en-ne/test-00000-of-00001.parquet b/en-ne/test-00000-of-00001.parquet
deleted file mode 100644
index 8848dd447394b1582988a753bfe84d9a2d3630a8..0000000000000000000000000000000000000000
--- a/en-ne/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2d2969355fa3c4e76a96e5015b43cd7b13e5518f7efaba0206d7a759b1548417
-size 93474
diff --git a/en-ne/train-00000-of-00001.parquet b/en-ne/train-00000-of-00001.parquet
deleted file mode 100644
index be1364f641000a879d5f8ca70f58f8bd8e6606f8..0000000000000000000000000000000000000000
--- a/en-ne/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:03d977b00ced9124dffc98b32d5aebf2ad1bf71cde13a4218f767af68caab606
-size 23913307
diff --git a/en-ne/validation-00000-of-00001.parquet b/en-ne/validation-00000-of-00001.parquet
deleted file mode 100644
index 65594a42f2fc67099645a9996fb1e61ef523b478..0000000000000000000000000000000000000000
--- a/en-ne/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4c8cd81e3c75ac9db98ca57e035e547c6b91b10091413eadc3270f59107dc540
-size 100742
diff --git a/en-nl/test-00000-of-00001.parquet b/en-nl/test-00000-of-00001.parquet
deleted file mode 100644
index 169bf8b34f6e06e227c5c2bd07c563e77a12aad9..0000000000000000000000000000000000000000
--- a/en-nl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8936fa98e90383303dbc558bbc13b8b6a4bb20068aa739fbf17a544d4a1297b5
-size 207972
diff --git a/en-nl/train-00000-of-00001.parquet b/en-nl/train-00000-of-00001.parquet
deleted file mode 100644
index 62f4fd5710bcdc6d6a20b3a6b6404f3b30a9502c..0000000000000000000000000000000000000000
--- a/en-nl/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ab9e66c872e117394aa986025bb8788e281a1301d71c23eb13ca42c9a7d0dce9
-size 82518283
diff --git a/en-nl/validation-00000-of-00001.parquet b/en-nl/validation-00000-of-00001.parquet
deleted file mode 100644
index e98e5789ccd4805d0d11f06ad9fe0f32f4222b38..0000000000000000000000000000000000000000
--- a/en-nl/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b975408ed783b895e0a800c24b4de0f75968fea6fae0f6e47347ad0587cd1a07
-size 197661
diff --git a/en-nn/test-00000-of-00001.parquet b/en-nn/test-00000-of-00001.parquet
deleted file mode 100644
index 23548001aeef7b7c662a8f5002ad7532364685c8..0000000000000000000000000000000000000000
--- a/en-nn/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ca52a5e3db6e1f88199ced5d849ce8e65d9bed4225b9f3a9dc25294c43e53130
-size 123205
diff --git a/en-nn/train-00000-of-00001.parquet b/en-nn/train-00000-of-00001.parquet
deleted file mode 100644
index db67310516da69a39b320a702d9aa5c1df57e535..0000000000000000000000000000000000000000
--- a/en-nn/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b9aac3317adc067ec928e31ab1590c6b4643b4ea06222a075dbccd7e7a689076
-size 24934219
diff --git a/en-nn/validation-00000-of-00001.parquet b/en-nn/validation-00000-of-00001.parquet
deleted file mode 100644
index ccb052e4e7fbbb8ea65e0f98887b26d587e60e62..0000000000000000000000000000000000000000
--- a/en-nn/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4a6e1995dce4e09524553a0dd54aa030d47cfb036a3285d6b3a65b2fabead1aa
-size 127252
diff --git a/en-no/test-00000-of-00001.parquet b/en-no/test-00000-of-00001.parquet
deleted file mode 100644
index 36187c466e839fe222cad6e3ffc6a65a4b2980d0..0000000000000000000000000000000000000000
--- a/en-no/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:535608ffe214ebb98b35261afdb5a029e24236f03c2c429801d4d281866cd6c1
-size 130584
diff --git a/en-no/train-00000-of-00001.parquet b/en-no/train-00000-of-00001.parquet
deleted file mode 100644
index 5484be85a9fda49c2f3973aed0e46ff4a142dee3..0000000000000000000000000000000000000000
--- a/en-no/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4396022d3406e94c7cfb3822d590fbf2bbbce4b1889364412b487c9edd62d6a7
-size 56013457
diff --git a/en-no/validation-00000-of-00001.parquet b/en-no/validation-00000-of-00001.parquet
deleted file mode 100644
index d2d746a095a193d370f2008a5e7296f32cb642d9..0000000000000000000000000000000000000000
--- a/en-no/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:33c06f8e19c7090c5b630216bf737d3a70ea5262ddc14fa0677da0d849c64396
-size 132959
diff --git a/en-oc/test-00000-of-00001.parquet b/en-oc/test-00000-of-00001.parquet
deleted file mode 100644
index 14d954705e3c806e08fa9007d87289c274fb29f6..0000000000000000000000000000000000000000
--- a/en-oc/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8165b771a044d61366832da6197a5aaaafec9f74735f310a077e2abc0145d5e3
-size 46938
diff --git a/en-oc/train-00000-of-00001.parquet b/en-oc/train-00000-of-00001.parquet
deleted file mode 100644
index 39106e3746d9e9807ff0cd0b1ad17ccc5eb09906..0000000000000000000000000000000000000000
--- a/en-oc/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c7f3565551ebe874312e0f8467ad702bf38ddef1eb8ec0985e8dba21e1455bcc
-size 1215954
diff --git a/en-oc/validation-00000-of-00001.parquet b/en-oc/validation-00000-of-00001.parquet
deleted file mode 100644
index 332a75d0f4be58b6178309f5bacd8cdd54523bd6..0000000000000000000000000000000000000000
--- a/en-oc/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3c7c87cdc32a289fef3158fb9ac721378a14ff6addaf1aa812690be1d062678f
-size 45446
diff --git a/en-or/test-00000-of-00001.parquet b/en-or/test-00000-of-00001.parquet
deleted file mode 100644
index 16cdaa907900e39c132d3d9867218e3024009e21..0000000000000000000000000000000000000000
--- a/en-or/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:42a741b4e09cae6cc1c542bebe69667110a66f7e54307b94057469935817d4ea
-size 86241
diff --git a/en-or/train-00000-of-00001.parquet b/en-or/train-00000-of-00001.parquet
deleted file mode 100644
index 02ee22ee40d638d790ff8a4ee0ee6c351d45c5d6..0000000000000000000000000000000000000000
--- a/en-or/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e8942ba631a84a84a32a5ac42cae2a404345276a07d54ab719ef4b2732160c08
-size 851790
diff --git a/en-or/validation-00000-of-00001.parquet b/en-or/validation-00000-of-00001.parquet
deleted file mode 100644
index 691c4eebf564e7dfc05ee21408948d56e0fdd88f..0000000000000000000000000000000000000000
--- a/en-or/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:39d573320ce6ebc65386b64649b7c3075b6b79ef0a52127558a04695dc6db2b9
-size 81940
diff --git a/en-pa/test-00000-of-00001.parquet b/en-pa/test-00000-of-00001.parquet
deleted file mode 100644
index ac0e54295addbd18422834d0b0d93e9099bc8c63..0000000000000000000000000000000000000000
--- a/en-pa/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:20dd645cab98f487f9ff60994e8d152722f113c1bfd8e1cd6002276e70e86d54
-size 69340
diff --git a/en-pa/train-00000-of-00001.parquet b/en-pa/train-00000-of-00001.parquet
deleted file mode 100644
index 597552f40adadc74136d4d4f242a707f563b3c91..0000000000000000000000000000000000000000
--- a/en-pa/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5a1430416131d0114e30bea3b678cc9c69235cfe88a9a6e59afc7f948ddabb5c
-size 5176352
diff --git a/en-pa/validation-00000-of-00001.parquet b/en-pa/validation-00000-of-00001.parquet
deleted file mode 100644
index f6e5d6126a784ead47429143f11e6acb7451fde3..0000000000000000000000000000000000000000
--- a/en-pa/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a3e8d454c5825821129eeb301040cc79165de41ad4b4c69e2b07069492978f3e
-size 69606
diff --git a/en-pl/test-00000-of-00001.parquet b/en-pl/test-00000-of-00001.parquet
deleted file mode 100644
index dac927b18505274a5b5dbd6d1728679893451560..0000000000000000000000000000000000000000
--- a/en-pl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:84214029f2545468f9c18cafbc361e11080b33153078a39242e2216794bf1208
-size 162455
diff --git a/en-pl/train-00000-of-00001.parquet b/en-pl/train-00000-of-00001.parquet
deleted file mode 100644
index 9682acd64dcd84c65591a07f2205b2bc76bea15c..0000000000000000000000000000000000000000
--- a/en-pl/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:55024ed840d6a8c99df6873e34673fc119de0066de7fea3317307767c54fc2f9
-size 73245510
diff --git a/en-pl/validation-00000-of-00001.parquet b/en-pl/validation-00000-of-00001.parquet
deleted file mode 100644
index f080a8778f2df2da7f652d389ea183a2cc275fc5..0000000000000000000000000000000000000000
--- a/en-pl/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a43034d96695de5b487cdc2d5f74ae8f3b5e911b5bc60b6910e687e28f4ad60a
-size 166079
diff --git a/en-ps/test-00000-of-00001.parquet b/en-ps/test-00000-of-00001.parquet
deleted file mode 100644
index 21ec4fcd70f3cd4ce41e7e783051c6fefa5373bc..0000000000000000000000000000000000000000
--- a/en-ps/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:63e461488b4beb902edea21a1de0634bfea3a6d8c81805a4bd2ea80f69f796fa
-size 45798
diff --git a/en-ps/train-00000-of-00001.parquet b/en-ps/train-00000-of-00001.parquet
deleted file mode 100644
index 97ef572cae53f6df2bd5da8ca0209bbb021c658d..0000000000000000000000000000000000000000
--- a/en-ps/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:27b71eb8d68edf2b4c6c66c7d32462d822c6c2a4ae5ca61e28df118803e04b4d
-size 2759492
diff --git a/en-ps/validation-00000-of-00001.parquet b/en-ps/validation-00000-of-00001.parquet
deleted file mode 100644
index 1d1273dc1f49c3b7eea523b6a75185bc53d72f78..0000000000000000000000000000000000000000
--- a/en-ps/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c226ed82aab57246bd9bb3f12f8af3143e91e02f40bd374f39a5257c308805b1
-size 46609
diff --git a/en-pt/test-00000-of-00001.parquet b/en-pt/test-00000-of-00001.parquet
deleted file mode 100644
index c95f3b4e4c7a4d4352ab4c5871d934ce20173e5a..0000000000000000000000000000000000000000
--- a/en-pt/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4181a70c9efea0f950afb62b863d3ae5170c17ed2b0289ea8c9a9e9898759ca3
-size 220098
diff --git a/en-pt/train-00000-of-00001.parquet b/en-pt/train-00000-of-00001.parquet
deleted file mode 100644
index c5a3e40cbd5c7720ce496952d0698c0cdba6af0e..0000000000000000000000000000000000000000
--- a/en-pt/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b02431a2521e2eb29e87bfea1dd3e1b3f94fb4875250d59d6ad81f303e54323d
-size 87224895
diff --git a/en-pt/validation-00000-of-00001.parquet b/en-pt/validation-00000-of-00001.parquet
deleted file mode 100644
index 5e703fb44400a25682fa7f4dcde6064166f7e9ad..0000000000000000000000000000000000000000
--- a/en-pt/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3a4a7f4070c3d1c517b83589500409e9221849c035bd109168ef3d37dd448054
-size 216914
diff --git a/en-ro/test-00000-of-00001.parquet b/en-ro/test-00000-of-00001.parquet
deleted file mode 100644
index dc83928b506c628ad080ecab96c15578f20a72bc..0000000000000000000000000000000000000000
--- a/en-ro/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:55a4d91c9484299b044b2a00eb6710adac52f6417f3d3c382556cbad6c837c5b
-size 152940
diff --git a/en-ro/train-00000-of-00001.parquet b/en-ro/train-00000-of-00001.parquet
deleted file mode 100644
index fb42edf3ea10ec58a8814979856967fff8d6bb65..0000000000000000000000000000000000000000
--- a/en-ro/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3121b23e07f008b3eec65bc5378baba6aab6b9f042c57906e7f30217b1f364fa
-size 65988584
diff --git a/en-ro/validation-00000-of-00001.parquet b/en-ro/validation-00000-of-00001.parquet
deleted file mode 100644
index 7f0f4e1b20ab46088a959762db87cee141ed5449..0000000000000000000000000000000000000000
--- a/en-ro/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8bb4fef306d597a9501d615fbab586debd031c210e5d0503557fd0385d4975b4
-size 152793
diff --git a/en-ru/test-00000-of-00001.parquet b/en-ru/test-00000-of-00001.parquet
deleted file mode 100644
index c83529f87b09990ade0f3d6c1de35d9270105d7f..0000000000000000000000000000000000000000
--- a/en-ru/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:96bf7751ebd69615e1377a06cf49bb7d2d153124c77764620de435d0afc71935
-size 310212
diff --git a/en-ru/train-00000-of-00001.parquet b/en-ru/train-00000-of-00001.parquet
deleted file mode 100644
index 24c7482db9568bdd718326a631f6d3814a1b48ab..0000000000000000000000000000000000000000
--- a/en-ru/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9b0b9580abef16fa28d74d2eeae359844e5ad372ebbfb6f8c8560add85aa69de
-size 123840514
diff --git a/en-ru/validation-00000-of-00001.parquet b/en-ru/validation-00000-of-00001.parquet
deleted file mode 100644
index 4e62d56b411c38a81cd3b37e4226ab9df37b87da..0000000000000000000000000000000000000000
--- a/en-ru/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0ecd5bc65260bc543a523470cf982a03e9a7109fb8d3c9d32f7eef97ae288504
-size 310090
diff --git a/en-rw/test-00000-of-00001.parquet b/en-rw/test-00000-of-00001.parquet
deleted file mode 100644
index d9deab377553275b7b78bfe57efcbcf7487bfd2c..0000000000000000000000000000000000000000
--- a/en-rw/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5e23d3780b31f3902276bc17ca1760bcf92b1c5c530f59995127678ebb84d0e3
-size 86126
diff --git a/en-rw/train-00000-of-00001.parquet b/en-rw/train-00000-of-00001.parquet
deleted file mode 100644
index 97d9912c286c60043b823319c0d3e3ad82028324..0000000000000000000000000000000000000000
--- a/en-rw/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c8c8714b96bff815c1f29b3ef5db88dec01d86789bc8db47af4a7de2f07b0d74
-size 9924064
diff --git a/en-rw/validation-00000-of-00001.parquet b/en-rw/validation-00000-of-00001.parquet
deleted file mode 100644
index 0c0accedad7bf304e4e77c457d2744529708dc48..0000000000000000000000000000000000000000
--- a/en-rw/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:68bea8776fe858b893275a40b4d97c68b16571edd27a64986487c88c689408cd
-size 83518
diff --git a/en-se/test-00000-of-00001.parquet b/en-se/test-00000-of-00001.parquet
deleted file mode 100644
index 7ce8efc227cd7d6281a0f04feb60720edc1ed439..0000000000000000000000000000000000000000
--- a/en-se/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d1acea79c0f209ace61dab651d61356aa90dbb91d520b7620f3c4f9f03a3608d
-size 53532
diff --git a/en-se/train-00000-of-00001.parquet b/en-se/train-00000-of-00001.parquet
deleted file mode 100644
index 8b778a971a79da3043a44de6081a5b33562555b7..0000000000000000000000000000000000000000
--- a/en-se/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:63960423a1220872e164e56ec585cf1d80a01566b1d70fbd035e71015db5bdf9
-size 1557607
diff --git a/en-se/validation-00000-of-00001.parquet b/en-se/validation-00000-of-00001.parquet
deleted file mode 100644
index 45caf38750d2b71b572c35eb9e8ab9168a1ee372..0000000000000000000000000000000000000000
--- a/en-se/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:bf168401560ecc706bf53ad9f8b1c746e9f1de8ba0462de29c219141ff673be1
-size 51706
diff --git a/en-sh/test-00000-of-00001.parquet b/en-sh/test-00000-of-00001.parquet
deleted file mode 100644
index 501ccc6ea7c67bd392a6136a41b874c91f77e1f9..0000000000000000000000000000000000000000
--- a/en-sh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ceaab3cdab67c47424d51a541e65604c8f9410bbd8f3d2690b8db398bd96a855
-size 370441
diff --git a/en-sh/train-00000-of-00001.parquet b/en-sh/train-00000-of-00001.parquet
deleted file mode 100644
index 0133adca2e399ed1422355f21c0cde278b428b13..0000000000000000000000000000000000000000
--- a/en-sh/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b24fe24663534f3e9d29273068117f01048a8132dde8b7b001504a7d47a88227
-size 39257987
diff --git a/en-sh/validation-00000-of-00001.parquet b/en-sh/validation-00000-of-00001.parquet
deleted file mode 100644
index 99cc12ae1fab5195894a37a345e945298ec0ac1f..0000000000000000000000000000000000000000
--- a/en-sh/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a28828f3280ae8867df9465d9f910f2a322da3d37eff123cd2362048eb2db19b
-size 360026
diff --git a/en-si/test-00000-of-00001.parquet b/en-si/test-00000-of-00001.parquet
deleted file mode 100644
index 2e2efaacada28e1e99ccb860e4c4ff77758e18fc..0000000000000000000000000000000000000000
--- a/en-si/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:94db26c3cec6a619fc53f0804e98152113ae9fe2ea1a308fae1a39d9fbad1784
-size 154795
diff --git a/en-si/train-00000-of-00001.parquet b/en-si/train-00000-of-00001.parquet
deleted file mode 100644
index 04f909bd1b77dc5ab882dc0ea1f381c6fbd13093..0000000000000000000000000000000000000000
--- a/en-si/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c2f1846d9eef839f5a93ac3a73fa4979e4d4d4006d0cd23ec9c1fee3b6b03baf
-size 65815918
diff --git a/en-si/validation-00000-of-00001.parquet b/en-si/validation-00000-of-00001.parquet
deleted file mode 100644
index 8974d54a868410951033bd4191f47866501165bd..0000000000000000000000000000000000000000
--- a/en-si/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ff8e8544bd254b09e0d34fcee7a220d1f5247c140c4c785ca86a8b3b690ca8fd
-size 153447
diff --git a/en-sk/test-00000-of-00001.parquet b/en-sk/test-00000-of-00001.parquet
deleted file mode 100644
index 2c5b433f022886c3e6cf457660bb4ceb7ab5e04b..0000000000000000000000000000000000000000
--- a/en-sk/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:30bf1829cafccb911eeee2e9fa7ed941ee0ae582e9f6b5b3bbf0367ccd685cb2
-size 195733
diff --git a/en-sk/train-00000-of-00001.parquet b/en-sk/train-00000-of-00001.parquet
deleted file mode 100644
index 505c2f04562ef2171ee96155ff8737a473a4ca9c..0000000000000000000000000000000000000000
--- a/en-sk/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3147898342822940b441d6f99e6ae4b3ca21a37f8a20834df6ee469ebd406e7e
-size 84833852
diff --git a/en-sk/validation-00000-of-00001.parquet b/en-sk/validation-00000-of-00001.parquet
deleted file mode 100644
index 1cb552df0713b80c2c18b844c86428bb0943d58d..0000000000000000000000000000000000000000
--- a/en-sk/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f730e756b3cec9c80189c37e9c58e974b13a3540ce12cc014857435fd8464055
-size 193745
diff --git a/en-sl/test-00000-of-00001.parquet b/en-sl/test-00000-of-00001.parquet
deleted file mode 100644
index 6d381fedbe44e2c36bb31a5de9894df4c5d128d0..0000000000000000000000000000000000000000
--- a/en-sl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f7bb94363a6f005cdbd98ca1271d456ad52c6e41980f92f9a56fe863b8aa58bc
-size 158451
diff --git a/en-sl/train-00000-of-00001.parquet b/en-sl/train-00000-of-00001.parquet
deleted file mode 100644
index 3266ebb1ceb8af4d87298b78331c68fe3e4478e1..0000000000000000000000000000000000000000
--- a/en-sl/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:075147b00f7d2eab253c7afc40bb4dabe48cb323dd9fa1594164dd5f1d8abb6c
-size 70395250
diff --git a/en-sl/validation-00000-of-00001.parquet b/en-sl/validation-00000-of-00001.parquet
deleted file mode 100644
index 614b9a61fb2d81c62efff09eaa7563c14a11bab4..0000000000000000000000000000000000000000
--- a/en-sl/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:33a811775f585b2e0dfca09fdaa1f59af80c18327565ab0e99f1ae52af6692bb
-size 154488
diff --git a/en-sq/test-00000-of-00001.parquet b/en-sq/test-00000-of-00001.parquet
deleted file mode 100644
index fea61a1e86ece4dc5f674f3ea5f12bcc6757fb18..0000000000000000000000000000000000000000
--- a/en-sq/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:32aa4b63a0fae6191dac244294b0d0ee7f85e2fd4114cabf0b7bf51aa5b92f44
-size 201712
diff --git a/en-sq/train-00000-of-00001.parquet b/en-sq/train-00000-of-00001.parquet
deleted file mode 100644
index f467a5ca8495cbbcd1d074385e3cac446619fa14..0000000000000000000000000000000000000000
--- a/en-sq/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:66f88dadaf983239cd4d9f0db3a8f764e1ffc4ca98dfb2985d65245eb34210f2
-size 78421684
diff --git a/en-sq/validation-00000-of-00001.parquet b/en-sq/validation-00000-of-00001.parquet
deleted file mode 100644
index 93307832711733f34a0927d53aa31fc223517e6e..0000000000000000000000000000000000000000
--- a/en-sq/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:eca01879295ff8d0c191b137b6c9a8492b733710146d5e59a19f8ce07f18c0d5
-size 194499
diff --git a/en-sr/test-00000-of-00001.parquet b/en-sr/test-00000-of-00001.parquet
deleted file mode 100644
index ca647bc3568278fa935beaccaa1ca9acfc0fc920..0000000000000000000000000000000000000000
--- a/en-sr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e79b64fe2ec33b00d64184d578384498e4cdad32f6c0f84f34a1e3aeaa0de24d
-size 141296
diff --git a/en-sr/train-00000-of-00001.parquet b/en-sr/train-00000-of-00001.parquet
deleted file mode 100644
index f296df2274a53719d6eba948c48b9758cacd14a1..0000000000000000000000000000000000000000
--- a/en-sr/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4b0e1d66a654df1981241e2f9aec6866ebdefa8273f181c121f86c677f926f29
-size 59977653
diff --git a/en-sr/validation-00000-of-00001.parquet b/en-sr/validation-00000-of-00001.parquet
deleted file mode 100644
index e3dc7b92eccd8493af4e730553b3c947845b3b04..0000000000000000000000000000000000000000
--- a/en-sr/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b39bae2da80ea0a9f1c0d969132ea0f1ccc5f430bf3b1bffb3c1e8126fee9f42
-size 144739
diff --git a/en-sv/test-00000-of-00001.parquet b/en-sv/test-00000-of-00001.parquet
deleted file mode 100644
index 14f2918784cd43b705bf49976854dbfb76cb5928..0000000000000000000000000000000000000000
--- a/en-sv/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:be1debb7f9d9e7c2966b5f563fbbb5911b12c2b4a2983ff174fef1cad9e6daaf
-size 193996
diff --git a/en-sv/train-00000-of-00001.parquet b/en-sv/train-00000-of-00001.parquet
deleted file mode 100644
index 706e5d40b818a3d145ef288499a96ddc542fe4c3..0000000000000000000000000000000000000000
--- a/en-sv/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:59b7e77af44a5b6d9f30cdd208823c5941a5f87efb46c4a50d7c83002936c8ad
-size 84636645
diff --git a/en-sv/validation-00000-of-00001.parquet b/en-sv/validation-00000-of-00001.parquet
deleted file mode 100644
index 97e15a835b7b11e46e1fdaba06c1f693e637c711..0000000000000000000000000000000000000000
--- a/en-sv/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2084554f862cc1a427fdb988d1c989be42a49589154a5e4b2040a354421a0615
-size 201486
diff --git a/en-ta/test-00000-of-00001.parquet b/en-ta/test-00000-of-00001.parquet
deleted file mode 100644
index 200bb922f6c55c2903a7f651e22beeb3f298d922..0000000000000000000000000000000000000000
--- a/en-ta/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7acbf1789b1ba29efd2c21e54cbc48c128050cca6f0750c6909731ba5cbfaee9
-size 164445
diff --git a/en-ta/train-00000-of-00001.parquet b/en-ta/train-00000-of-00001.parquet
deleted file mode 100644
index 2f0cd58bd5e42a5288720ccbe0dafc10f106545d..0000000000000000000000000000000000000000
--- a/en-ta/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b8287ee0c6f4cda982020bee9f9aab554c3a132ec29a06bdc3815525d0ccd900
-size 33319633
diff --git a/en-ta/validation-00000-of-00001.parquet b/en-ta/validation-00000-of-00001.parquet
deleted file mode 100644
index 97832ea6500f3ea38a11aed5ae24ff492cb52bd0..0000000000000000000000000000000000000000
--- a/en-ta/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:571075ba1ed0ce0cb367f707b4d831cd2f162cfb4cfde49f12e573a72050b706
-size 158616
diff --git a/en-te/test-00000-of-00001.parquet b/en-te/test-00000-of-00001.parquet
deleted file mode 100644
index 790783c4a70093dec265ab763c70c6020dedf3ef..0000000000000000000000000000000000000000
--- a/en-te/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1f4a183f564300d34906f91351fb1e08408dff964282f1d16b723f418f09b0e9
-size 99554
diff --git a/en-te/train-00000-of-00001.parquet b/en-te/train-00000-of-00001.parquet
deleted file mode 100644
index 422d5af14b0e3b9617aa77e06cbacd951940c420..0000000000000000000000000000000000000000
--- a/en-te/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ac896be7ec2064cbd9c9d00a7943a4ba35e2c90a8af5d7390e21a204a1132ffe
-size 3847809
diff --git a/en-te/validation-00000-of-00001.parquet b/en-te/validation-00000-of-00001.parquet
deleted file mode 100644
index 8d4e372c2f8cf208a5569e8f38a34150f1508e5b..0000000000000000000000000000000000000000
--- a/en-te/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8a70a41218f2b8f318f4a62ab9e6031131a66ed4aa2ff4dde7f24d19bb4630cf
-size 100304
diff --git a/en-tg/test-00000-of-00001.parquet b/en-tg/test-00000-of-00001.parquet
deleted file mode 100644
index 7d6f7e11d2c5d4d9424b67e0600b264acaeb8190..0000000000000000000000000000000000000000
--- a/en-tg/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0c496ff58db39570432e2bd14e233d369f633d689722b777a0f473db2a8258d0
-size 210080
diff --git a/en-tg/train-00000-of-00001.parquet b/en-tg/train-00000-of-00001.parquet
deleted file mode 100644
index 747e987e9629e81e912db789a43442069010e585..0000000000000000000000000000000000000000
--- a/en-tg/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b71094c9c43fed46ec9ffdbb55fb7fbeca1bb2d136638c61f4e0745b5e4fbe03
-size 20822821
diff --git a/en-tg/validation-00000-of-00001.parquet b/en-tg/validation-00000-of-00001.parquet
deleted file mode 100644
index d79555d10904c7dba2dd4f65a906a555ce857d87..0000000000000000000000000000000000000000
--- a/en-tg/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:178f322c2df8c1e4d172f90d22fb71dc246a4672166f7aec65779865847702fb
-size 209767
diff --git a/en-th/test-00000-of-00001.parquet b/en-th/test-00000-of-00001.parquet
deleted file mode 100644
index 221efb6121578d3be9b8fde677d0ee0fa90214a9..0000000000000000000000000000000000000000
--- a/en-th/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b64f10497015cdbd993436d9de9c15a9ce72b422588bb20153b6133353572f6a
-size 164056
diff --git a/en-th/train-00000-of-00001.parquet b/en-th/train-00000-of-00001.parquet
deleted file mode 100644
index aea6eab08536277538d2939bb6114f8d9a471458..0000000000000000000000000000000000000000
--- a/en-th/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5c1f7a1c664e3c8975360c7d0849841921564617d804e1acf757f1141642fbdb
-size 75214489
diff --git a/en-th/validation-00000-of-00001.parquet b/en-th/validation-00000-of-00001.parquet
deleted file mode 100644
index e65d6a63238bb9e45990513d231635273f66eced..0000000000000000000000000000000000000000
--- a/en-th/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d153e63520e03eb1bd9bc6352b791701747f1bef16f9ce76ed055246e220386d
-size 161442
diff --git a/en-tk/test-00000-of-00001.parquet b/en-tk/test-00000-of-00001.parquet
deleted file mode 100644
index c64b6820f86ed4e33e06ec3abc7ba3d7d7d42ba5..0000000000000000000000000000000000000000
--- a/en-tk/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7bc37862e3376a4425cb3806fc1c635cda5f01119789348fbcc38dbad0e69eb0
-size 27223
diff --git a/en-tk/train-00000-of-00001.parquet b/en-tk/train-00000-of-00001.parquet
deleted file mode 100644
index b0af2e7958451a034a0febc057df212c16fa6078..0000000000000000000000000000000000000000
--- a/en-tk/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:53e4232ae94a8b9985f9a6a8a18ff60ce6fada83416b97aeab9b6a8296fb2120
-size 363794
diff --git a/en-tk/validation-00000-of-00001.parquet b/en-tk/validation-00000-of-00001.parquet
deleted file mode 100644
index 3d3d4b11e6c6f3b7908171e1ec8559b9270894e4..0000000000000000000000000000000000000000
--- a/en-tk/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a817442f3f6e1a966aa682b297d13acee07b537620583c8845211174d6cdb44d
-size 26739
diff --git a/en-tr/test-00000-of-00001.parquet b/en-tr/test-00000-of-00001.parquet
deleted file mode 100644
index b3ec1cdd1c2d35a0668fa70354abffed538e93cf..0000000000000000000000000000000000000000
--- a/en-tr/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a512f2d6fa5a215eeeda862fff9909b4d5b048219055f11bdef0892f76d8b143
-size 138865
diff --git a/en-tr/train-00000-of-00001.parquet b/en-tr/train-00000-of-00001.parquet
deleted file mode 100644
index e5cff7e8ad27a69efcec9e171d8a544c8e30f598..0000000000000000000000000000000000000000
--- a/en-tr/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a1a2979fe986891682913dba5da0888323a5d21b3b3249f69c373c9fb3a3b5ae
-size 60088392
diff --git a/en-tr/validation-00000-of-00001.parquet b/en-tr/validation-00000-of-00001.parquet
deleted file mode 100644
index 225545fa57544ecca6c3975eb7f6e75f9ce8feea..0000000000000000000000000000000000000000
--- a/en-tr/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0c52bc5106b2e4f84ceffd79a15e913f8d4fdfb3eacc5377804c6e668f243a48
-size 137664
diff --git a/en-tt/test-00000-of-00001.parquet b/en-tt/test-00000-of-00001.parquet
deleted file mode 100644
index 7f97e3db987fc1617bd987d154ff337d66b7e6ed..0000000000000000000000000000000000000000
--- a/en-tt/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9566eace863edc55cc8419ddf35cefa519380040e4bdc3be560f2d8376dd3984
-size 342317
diff --git a/en-tt/train-00000-of-00001.parquet b/en-tt/train-00000-of-00001.parquet
deleted file mode 100644
index ed8b0720a2bd708411b86318435ed1d0ff5bdd0e..0000000000000000000000000000000000000000
--- a/en-tt/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4bb4552d40677b7d57866224226b42d21c282196d38db7680812475ac13e61ba
-size 18095148
diff --git a/en-tt/validation-00000-of-00001.parquet b/en-tt/validation-00000-of-00001.parquet
deleted file mode 100644
index 037b974f609b020dc836881770055929c6e17609..0000000000000000000000000000000000000000
--- a/en-tt/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:28747f764f1e7bf37711251aed2ef0fbf415c1fa264ce8f15d7b41cace78e9e7
-size 349533
diff --git a/en-ug/test-00000-of-00001.parquet b/en-ug/test-00000-of-00001.parquet
deleted file mode 100644
index 34ab59c091d302525c9413d6f9c604f4697a97ff..0000000000000000000000000000000000000000
--- a/en-ug/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:175f5c57b8e5573c48b674dd54023f38ce683916cb317ca2b28e2aca13056de1
-size 323589
diff --git a/en-ug/train-00000-of-00001.parquet b/en-ug/train-00000-of-00001.parquet
deleted file mode 100644
index b43c5e8c5797ef86adc6ad6231e13751cbe2aa00..0000000000000000000000000000000000000000
--- a/en-ug/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:96366787fd3f13c3e99c7a6ae0c4f2a7c809ac993524b4559e6f1587a967525c
-size 15362122
diff --git a/en-ug/validation-00000-of-00001.parquet b/en-ug/validation-00000-of-00001.parquet
deleted file mode 100644
index 9d3017c1db329ad2ea63b0b9afde918c38afff01..0000000000000000000000000000000000000000
--- a/en-ug/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:978db16240a97e4caa22cda5e673f26b3202bc4f867dd58cdc0e8d7ef941aa5d
-size 325661
diff --git a/en-uk/test-00000-of-00001.parquet b/en-uk/test-00000-of-00001.parquet
deleted file mode 100644
index 5df11cd79f54172b967c06e711a195bc220806db..0000000000000000000000000000000000000000
--- a/en-uk/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2a6743962f8e16fd8bf71c34ebd242bf18d5d8eeb0e9724cf4d6f878028a7019
-size 168266
diff --git a/en-uk/train-00000-of-00001.parquet b/en-uk/train-00000-of-00001.parquet
deleted file mode 100644
index 8df1b554ab6398c032b9dac6921ce9d8b7ac670d..0000000000000000000000000000000000000000
--- a/en-uk/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9cbc4eae1002229b7152d51042a074771455ab1c006087382ce3daa351fa41ff
-size 70819354
diff --git a/en-uk/validation-00000-of-00001.parquet b/en-uk/validation-00000-of-00001.parquet
deleted file mode 100644
index 79a7706629a0a80ad73e39d69dca9bd9abf86771..0000000000000000000000000000000000000000
--- a/en-uk/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:be8fcf96bd7cfd4e31fcba88ce6677a9b832273ed1d6abb29b9a313245dbfdc9
-size 168062
diff --git a/en-ur/test-00000-of-00001.parquet b/en-ur/test-00000-of-00001.parquet
deleted file mode 100644
index 98c3cb919ff8b4051617facdb4ec418fa1c4d007..0000000000000000000000000000000000000000
--- a/en-ur/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:60552603e9768fd43cc5bae6652cff0a4a4863ef21894d02129fcdfbaa1a2930
-size 300728
diff --git a/en-ur/train-00000-of-00001.parquet b/en-ur/train-00000-of-00001.parquet
deleted file mode 100644
index 60b6d47d72d7b65abd7a52d272cdac175a4303b1..0000000000000000000000000000000000000000
--- a/en-ur/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c7158d1f70ecdcdc04b1bdc355d4f9f57149eae188f9213727e514f02b3d29d6
-size 147739018
diff --git a/en-ur/validation-00000-of-00001.parquet b/en-ur/validation-00000-of-00001.parquet
deleted file mode 100644
index 27f009d99fac456179cfd07a73c6bbef47971be4..0000000000000000000000000000000000000000
--- a/en-ur/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:eabed2146ca4f0e2bdd7f07102f8922a9be68399c84d98c77efdf5a6c381f381
-size 296298
diff --git a/en-uz/test-00000-of-00001.parquet b/en-uz/test-00000-of-00001.parquet
deleted file mode 100644
index 4af5d968d1ccb8cf5530c3b24617ccc6f377b7ed..0000000000000000000000000000000000000000
--- a/en-uz/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cc3f5df46d6876de2be4cbd9322bfb052c65d0a586a6adf1bea6fa4ca01db4ea
-size 221531
diff --git a/en-uz/train-00000-of-00001.parquet b/en-uz/train-00000-of-00001.parquet
deleted file mode 100644
index 5a474cf23c61ef29bee350648eebb2159ebe8b02..0000000000000000000000000000000000000000
--- a/en-uz/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:97ec86928425abc9423b881ee7c785cff83fc8a2a8f33643e3e4bebf01c3835a
-size 21435777
diff --git a/en-uz/validation-00000-of-00001.parquet b/en-uz/validation-00000-of-00001.parquet
deleted file mode 100644
index caab762af4a21608f423695c790ebba65c1430ea..0000000000000000000000000000000000000000
--- a/en-uz/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:409ff2719246b5ff96a1d1934e820eeaa9c24bb09155cba6715e985e451ecc92
-size 216228
diff --git a/en-vi/test-00000-of-00001.parquet b/en-vi/test-00000-of-00001.parquet
deleted file mode 100644
index 67ff3891213e0faf6af267412519f2ee24fe9ce1..0000000000000000000000000000000000000000
--- a/en-vi/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f53af271885d858845bdfc9b8de5477d85cb7d3211dde5b4c9078a3c192b8f03
-size 137302
diff --git a/en-vi/train-00000-of-00001.parquet b/en-vi/train-00000-of-00001.parquet
deleted file mode 100644
index 1b3bf0141a94803c345953fe1f34b26af83afbc9..0000000000000000000000000000000000000000
--- a/en-vi/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:27537bc369893219b0ac73ca56d8d1e36762fb34a7988f7b96e5c94c4b792b9d
-size 58975355
diff --git a/en-vi/validation-00000-of-00001.parquet b/en-vi/validation-00000-of-00001.parquet
deleted file mode 100644
index fd1002eb8bdbea26a5c357f5ce15f01f5d56a871..0000000000000000000000000000000000000000
--- a/en-vi/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:14cb5358d8a3583c645dfa0269255127d4c4e79833d4fbbbdabc52038f824226
-size 138195
diff --git a/en-wa/test-00000-of-00001.parquet b/en-wa/test-00000-of-00001.parquet
deleted file mode 100644
index 5039af1ec675ab89205f4a078f5110ff86ebfb57..0000000000000000000000000000000000000000
--- a/en-wa/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:febf430d01e333bce4e394d4688e75fcd0c7afb503b82ce089c7ebe63baafa7c
-size 50627
diff --git a/en-wa/train-00000-of-00001.parquet b/en-wa/train-00000-of-00001.parquet
deleted file mode 100644
index 3e22596bd5f8bbf0add4bb9d8658ecab41a586b4..0000000000000000000000000000000000000000
--- a/en-wa/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:87c677a98824ebecc1eac06146ec1527c2f960e64f6a82083cef7a2aece4fc73
-size 4411475
diff --git a/en-wa/validation-00000-of-00001.parquet b/en-wa/validation-00000-of-00001.parquet
deleted file mode 100644
index 6e40ce607c8e454a90e47cea1a887a34d68c5c52..0000000000000000000000000000000000000000
--- a/en-wa/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9186e7b5b1332a6d44cff1a7d19d83390c740f7134dd8d0b614e7172252478cd
-size 50102
diff --git a/en-xh/test-00000-of-00001.parquet b/en-xh/test-00000-of-00001.parquet
deleted file mode 100644
index 8dfb90c707813f0fb4fb08ae3974fc355fbc4a06..0000000000000000000000000000000000000000
--- a/en-xh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b83bb2198f91dcfa94f5ff03cec898f35b7f9f48b1ebcd529562eb31f86d6248
-size 229058
diff --git a/en-xh/train-00000-of-00001.parquet b/en-xh/train-00000-of-00001.parquet
deleted file mode 100644
index d9c26a710699ceb728a99342b40a3f21319acf16..0000000000000000000000000000000000000000
--- a/en-xh/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a560eff7635b3ac06b5106a1940973abfa77c692c20bd8f99cca74b517352566
-size 37064235
diff --git a/en-xh/validation-00000-of-00001.parquet b/en-xh/validation-00000-of-00001.parquet
deleted file mode 100644
index bf72ea1736c3c1c9dd966a278bcba0d7d4b9b29e..0000000000000000000000000000000000000000
--- a/en-xh/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b7b0944fb48c947f44fd9958f893c69eeb429ce424d2f9dad48605f85564a5dd
-size 226072
diff --git a/en-yi/test-00000-of-00001.parquet b/en-yi/test-00000-of-00001.parquet
deleted file mode 100644
index fb4947562f9cdfd2f04bd8a12413d4409b066a02..0000000000000000000000000000000000000000
--- a/en-yi/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b52e02c7bf2763b1bf2bb40455d1bae4eb8fa51a5df1a9d1bfba5208d8211541
-size 31484
diff --git a/en-yi/train-00000-of-00001.parquet b/en-yi/train-00000-of-00001.parquet
deleted file mode 100644
index ae2df695eccf82df9f0aaac679084c36088b19d1..0000000000000000000000000000000000000000
--- a/en-yi/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3975470544c8e5b6013bece8db85a1a1418ce10161d8b84f47779ff6e49a8fd3
-size 586193
diff --git a/en-yi/validation-00000-of-00001.parquet b/en-yi/validation-00000-of-00001.parquet
deleted file mode 100644
index ece2c2840c56ac61d720f7035e09c6c188c5b73c..0000000000000000000000000000000000000000
--- a/en-yi/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b1d3e1cb8c69f0ed6ba76aefec9abb37e14305425a369b082a17da3e324c7d27
-size 32853
diff --git a/en-yo/train-00000-of-00001.parquet b/en-yo/train-00000-of-00001.parquet
deleted file mode 100644
index 9932bd31e752074c01f647a25fec5c53adf47d4f..0000000000000000000000000000000000000000
--- a/en-yo/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:dcca98c6ff0e081f4440506b259400f2f436fb904b6ae771a5021f63942f1617
-size 391299
diff --git a/en-zh/test-00000-of-00001.parquet b/en-zh/test-00000-of-00001.parquet
deleted file mode 100644
index b297aebfcad9b39c17b1e1c5bea7fd6b41db4756..0000000000000000000000000000000000000000
--- a/en-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:571b811b20859d801561f62a21cc0e8a2fb034dd54bc960bb4b879dc6345fd19
-size 354971
diff --git a/en-zh/train-00000-of-00001.parquet b/en-zh/train-00000-of-00001.parquet
deleted file mode 100644
index 3cf957965a27d7b1108fc37fede220c2b1760814..0000000000000000000000000000000000000000
--- a/en-zh/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e48b522a92493526c39bb9417a5ece77cf406b07c59797bc652072b296914dd9
-size 142700288
diff --git a/en-zh/validation-00000-of-00001.parquet b/en-zh/validation-00000-of-00001.parquet
deleted file mode 100644
index 8405ee95e2efacb33cbda9cdd477ac4d41a18b8f..0000000000000000000000000000000000000000
--- a/en-zh/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ce0428a37c741a3c55ae707ca7b12fc0410b015212ad604f9b305f810188434c
-size 359497
diff --git a/en-zu/test-00000-of-00001.parquet b/en-zu/test-00000-of-00001.parquet
deleted file mode 100644
index c6de85500484277ea0a3d5f10661ae7b20add78e..0000000000000000000000000000000000000000
--- a/en-zu/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:96c0eb1270d719e6717f18dfa75a94fea1f314d5456adbec4f60717d4e1f220e
-size 58107
diff --git a/en-zu/train-00000-of-00001.parquet b/en-zu/train-00000-of-00001.parquet
deleted file mode 100644
index e75984fdc6168d3a3245735a568b6d9707cb3482..0000000000000000000000000000000000000000
--- a/en-zu/train-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:067d06f0cf7e57fbe876441cd89b25443b8ad52db0888b238845989298a34b16
-size 1801447
diff --git a/en-zu/validation-00000-of-00001.parquet b/en-zu/validation-00000-of-00001.parquet
deleted file mode 100644
index b9e8b65f6c03063ef58b93add160055dd5274d8f..0000000000000000000000000000000000000000
--- a/en-zu/validation-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:578ba9f096ab81b7e99a02cd6f02e1ee14f26351d09c762dd129242bb0cdc55a
-size 58889
diff --git a/fr-nl/test-00000-of-00001.parquet b/fr-nl/test-00000-of-00001.parquet
deleted file mode 100644
index 31beebddb0eb83d92e43a8155ab13181d5157021..0000000000000000000000000000000000000000
--- a/fr-nl/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4abce043f64de00571b83910fdd47c621fdf1d4fafe339140a449f28bd7b4856
-size 261290
diff --git a/fr-ru/test-00000-of-00001.parquet b/fr-ru/test-00000-of-00001.parquet
deleted file mode 100644
index 7b54b8d4fe1a21c40a59315caca01e8085284566..0000000000000000000000000000000000000000
--- a/fr-ru/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b7ddcc16837727a54fb245b3d8ad4553296d269160d87df231ed06c4c73c7b6b
-size 426179
diff --git a/fr-zh/test-00000-of-00001.parquet b/fr-zh/test-00000-of-00001.parquet
deleted file mode 100644
index 82b4b27f4a9ee113301f28712103232125f4766c..0000000000000000000000000000000000000000
--- a/fr-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d51908bc37b7a31919de9e5c7033ea3e560a99c6e9d983ca9543c46a01a38206
-size 418661
diff --git a/nl-ru/test-00000-of-00001.parquet b/nl-ru/test-00000-of-00001.parquet
deleted file mode 100644
index 7dd34b0dede35e5227bb767888e648d1feac71f7..0000000000000000000000000000000000000000
--- a/nl-ru/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:55ea3de5c476c02eb73037d6d4db6c207e68c65c74c54a0d75e2f7d9a223334c
-size 168666
diff --git a/nl-zh/test-00000-of-00001.parquet b/nl-zh/test-00000-of-00001.parquet
deleted file mode 100644
index 92a324312a3a637476392bca680b121e91ce3695..0000000000000000000000000000000000000000
--- a/nl-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:38dc2965e7933bfe4c3bab877d92543c90a4af67c8973beccd261eef65be48a6
-size 146191
diff --git a/opus100.py b/opus100.py
new file mode 100644
index 0000000000000000000000000000000000000000..dbc9e5ff817cefc58ee95e24da76027dade65254
--- /dev/null
+++ b/opus100.py
@@ -0,0 +1,271 @@
+# coding=utf-8
+# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""OPUS-100"""
+
+
+import datasets
+
+
+_CITATION = """\
+@misc{zhang2020improving,
+      title={Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation},
+      author={Biao Zhang and Philip Williams and Ivan Titov and Rico Sennrich},
+      year={2020},
+      eprint={2004.11867},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
+}
+"""
+
+_DESCRIPTION = """\
+OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side.
+The corpus covers 100 languages (including English).OPUS-100 contains approximately 55M sentence pairs.
+Of the 99 language pairs, 44 have 1M sentence pairs of training data, 73 have at least 100k, and 95 have at least 10k.
+"""
+
+_URL = {
+    "supervised": "https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-{}-v1.0.tar.gz",
+    "zero-shot": "https://object.pouta.csc.fi/OPUS-100/v1.0/opus-100-corpus-zeroshot-v1.0.tar.gz",
+}
+
+_SupervisedLanguagePairs = [
+    "af-en",
+    "am-en",
+    "an-en",
+    "ar-en",
+    "as-en",
+    "az-en",
+    "be-en",
+    "bg-en",
+    "bn-en",
+    "br-en",
+    "bs-en",
+    "ca-en",
+    "cs-en",
+    "cy-en",
+    "da-en",
+    "de-en",
+    "dz-en",
+    "el-en",
+    "en-eo",
+    "en-es",
+    "en-et",
+    "en-eu",
+    "en-fa",
+    "en-fi",
+    "en-fr",
+    "en-fy",
+    "en-ga",
+    "en-gd",
+    "en-gl",
+    "en-gu",
+    "en-ha",
+    "en-he",
+    "en-hi",
+    "en-hr",
+    "en-hu",
+    "en-hy",
+    "en-id",
+    "en-ig",
+    "en-is",
+    "en-it",
+    "en-ja",
+    "en-ka",
+    "en-kk",
+    "en-km",
+    "en-ko",
+    "en-kn",
+    "en-ku",
+    "en-ky",
+    "en-li",
+    "en-lt",
+    "en-lv",
+    "en-mg",
+    "en-mk",
+    "en-ml",
+    "en-mn",
+    "en-mr",
+    "en-ms",
+    "en-mt",
+    "en-my",
+    "en-nb",
+    "en-ne",
+    "en-nl",
+    "en-nn",
+    "en-no",
+    "en-oc",
+    "en-or",
+    "en-pa",
+    "en-pl",
+    "en-ps",
+    "en-pt",
+    "en-ro",
+    "en-ru",
+    "en-rw",
+    "en-se",
+    "en-sh",
+    "en-si",
+    "en-sk",
+    "en-sl",
+    "en-sq",
+    "en-sr",
+    "en-sv",
+    "en-ta",
+    "en-te",
+    "en-tg",
+    "en-th",
+    "en-tk",
+    "en-tr",
+    "en-tt",
+    "en-ug",
+    "en-uk",
+    "en-ur",
+    "en-uz",
+    "en-vi",
+    "en-wa",
+    "en-xh",
+    "en-yi",
+    "en-yo",
+    "en-zh",
+    "en-zu",
+]
+
+_0shotLanguagePairs = [
+    "ar-de",
+    "ar-fr",
+    "ar-nl",
+    "ar-ru",
+    "ar-zh",
+    "de-fr",
+    "de-nl",
+    "de-ru",
+    "de-zh",
+    "fr-nl",
+    "fr-ru",
+    "fr-zh",
+    "nl-ru",
+    "nl-zh",
+    "ru-zh",
+]
+
+
+class Opus100Config(datasets.BuilderConfig):
+    """BuilderConfig for Opus100"""
+
+    def __init__(self, language_pair, **kwargs):
+        super().__init__(**kwargs)
+        """
+
+        Args:
+            language_pair: language pair, you want to load
+            **kwargs: keyword arguments forwarded to super.
+        """
+        self.language_pair = language_pair
+
+
+class Opus100(datasets.GeneratorBasedBuilder):
+    """OPUS-100 is English-centric, meaning that all training pairs include English on either the source or target side."""
+
+    VERSION = datasets.Version("1.0.0")
+
+    BUILDER_CONFIG_CLASS = Opus100Config
+    BUILDER_CONFIGS = [
+        Opus100Config(name=pair, description=_DESCRIPTION, language_pair=pair)
+        for pair in _SupervisedLanguagePairs + _0shotLanguagePairs
+    ]
+
+    def _info(self):
+        src_tag, tgt_tag = self.config.language_pair.split("-")
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features({"translation": datasets.features.Translation(languages=(src_tag, tgt_tag))}),
+            supervised_keys=(src_tag, tgt_tag),
+            homepage="http://opus.nlpl.eu/opus-100.php",
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+
+        lang_pair = self.config.language_pair
+        src_tag, tgt_tag = lang_pair.split("-")
+
+        domain = "supervised"
+        if lang_pair in _0shotLanguagePairs:
+            domain = "zero-shot"
+
+        if domain == "supervised":
+            archive = dl_manager.download(_URL["supervised"].format(lang_pair))
+        elif domain == "zero-shot":
+            archive = dl_manager.download(_URL["zero-shot"])
+
+        data_dir = "/".join(["opus-100-corpus", "v1.0", domain, lang_pair])
+        output = []
+
+        test = datasets.SplitGenerator(
+            name=datasets.Split.TEST,
+            # These kwargs will be passed to _generate_examples
+            gen_kwargs={
+                "filepath": f"{data_dir}/opus.{lang_pair}-test.{src_tag}",
+                "labelpath": f"{data_dir}/opus.{lang_pair}-test.{tgt_tag}",
+                "files": dl_manager.iter_archive(archive),
+            },
+        )
+
+        available_files = [path for path, _ in dl_manager.iter_archive(archive)]
+        if f"{data_dir}/opus.{lang_pair}-test.{src_tag}" in available_files:
+            output.append(test)
+
+        if domain == "supervised":
+
+            train = datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": f"{data_dir}/opus.{lang_pair}-train.{src_tag}",
+                    "labelpath": f"{data_dir}/opus.{lang_pair}-train.{tgt_tag}",
+                    "files": dl_manager.iter_archive(archive),
+                },
+            )
+
+            if f"{data_dir}/opus.{lang_pair}-train.{src_tag}" in available_files:
+                output.append(train)
+
+            valid = datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={
+                    "filepath": f"{data_dir}/opus.{lang_pair}-dev.{src_tag}",
+                    "labelpath": f"{data_dir}/opus.{lang_pair}-dev.{tgt_tag}",
+                    "files": dl_manager.iter_archive(archive),
+                },
+            )
+
+            if f"{data_dir}/opus.{lang_pair}-dev.{src_tag}" in available_files:
+                output.append(valid)
+
+        return output
+
+    def _generate_examples(self, filepath, labelpath, files):
+        """Yields examples."""
+        src_tag, tgt_tag = self.config.language_pair.split("-")
+        src, tgt = None, None
+        for path, f in files:
+            if path == filepath:
+                src = f.read().decode("utf-8").split("\n")[:-1]
+            elif path == labelpath:
+                tgt = f.read().decode("utf-8").split("\n")[:-1]
+            if src is not None and tgt is not None:
+                for idx, (s, t) in enumerate(zip(src, tgt)):
+                    yield idx, {"translation": {src_tag: s, tgt_tag: t}}
+                break
diff --git a/ru-zh/test-00000-of-00001.parquet b/ru-zh/test-00000-of-00001.parquet
deleted file mode 100644
index a0fd243b7fb9a0661a70c2f2b5b2c13fd538879b..0000000000000000000000000000000000000000
--- a/ru-zh/test-00000-of-00001.parquet
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6d2c2e22b40d27227ca3f96073f723f4f3fa5a8a259f4f38f14154274e6d2a2a
-size 534430