Update files from the datasets library (from 1.6.0)

Release notes: https://github.com/huggingface/datasets/releases/tag/1.6.0

dummy/common_gen/{1.0.0 → 1.1.0}/dummy_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59e6e129316e63c1a54d6ef7bd9a0c540c62ff4c1a36df33327a9e3facd3b4e3
-size 2333
+oid sha256:686315482fae8bbd0d847372d76c07aa9119c374ab780aaed5b9f41979349a92
+size 4735

dummy/cs_restaurants/{1.0.0 → 1.1.0}/dummy_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25cab7c6d4e34d5fb08d476ce278d5e01ee6211c1b99629795bef0156e6aa785
-size 1841
+oid sha256:0fbd30e7d0d1d211ea2da944c800ec76b8bdeebebc4f696b792237069e8ae1d9
+size 4230

dummy/e2e_nlg/{1.0.0 → 1.1.0}/dummy_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02c353e0079daa9fc1b1d2e63f4f94a389ab340ad15f78e874c226dc355836ae
-size 1338
+oid sha256:2d266d483e50599c7b4eedce57d5df1f92f000aa90cbd7fa31eb57f7959a94f1
+size 3689

dummy/mlsum_de/{1.0.0 → 1.1.0}/dummy_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:66edd328604c4aa3a9967d81537b438d22202f9830dbda85056fd561b3ae6d4b
-size 17325
+oid sha256:df4ed9c1975aff72e507da3a8edc236321945612a835a54ca93b5ea2ed0d4c61
+size 34048

dummy/mlsum_es/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:587408dc43119abcf6d3000266a916233ac5ccabfb5f01e87da55539df303597
-size 23066

dummy/mlsum_es/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:816565f2e923373a93c639d308ab17ca2faae27d226c8186cb391e22db46bc36
+size 40918

dummy/schema_guided_dialog/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b27ba315658a1cabdfc16ef83eee6bc525347183906bad9dddf4d33b5c48c11a
+size 12875

dummy/totto/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a730949a9fa8a9d5affcd9ec6069470a531903856f97f73971d5a3ef2f8a8801
-size 24427

dummy/totto/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3aa6b2a296ad9a2c6f52066352132f297fd0eb833c106fbd76fe387c6772a19
+size 32908

dummy/web_nlg_en/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:11e43d5dc953eae0070317b95ad533a46b8f2dc0c5751d33234d29b1e832bc75
-size 2623

dummy/{schema_guided_dialog/1.0.0 → web_nlg_en/1.1.0}/dummy_data.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:73573be9eb634941d2daa888cfcf504cc3bbabab7a8e0d1712a55e7037b230b0
-size 1899
+oid sha256:ca1d503751ebb251b1b9315e03d222ba85a6f70d69a80c42259ed0b83a307854
+size 5754

dummy/web_nlg_ru/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:428efef997ade4b3c7f9b110a681d2a24abe57f40c4f342826f57f85f8fb9ca7
-size 3822

dummy/web_nlg_ru/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64caee03808e0724f6abe03cd8d438305520b99a8d4c8016b3757ed9d40ac5e4
+size 6279

dummy/wiki_auto_asset_turk/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:80352624751ac6f5a3cb44439470ec3ffec0a901e9eafe83bcf14c61372dbfa0
-size 10318

dummy/wiki_auto_asset_turk/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1344de60da0c4ca84f918e8c587d3fed5326a1deed5924566efe9525f7645843
+size 23815

dummy/wiki_lingua_arabic_ar/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9bb7f2baf7423770d9f44d84084850c23e36cbf6462b94e5943a49a35d29282
+size 17747

dummy/wiki_lingua_chinese_zh/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5647262bf23f33dcc884c5674b5f43eca71fc25bddbb9eed291efc9feb7bf05c
+size 18261

dummy/wiki_lingua_czech_cs/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e829391b38736a189bcaff05356983c52d500fca4bd86b186b26501989e260dd
+size 21235

dummy/wiki_lingua_dutch_nl/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b567d06578f0a7793a7435058601533b4d279ed9a86879fe7eaa76ed048157e
+size 17063

dummy/wiki_lingua_english_en/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:472a8592c0bf412172670a1fafd23a54e4bb42ab58c45fae69927420db31a4d5
+size 9106

dummy/wiki_lingua_french_fr/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b23ebb87a54b58bfea9ac012e6a894f381ae560df51218f25f2fe6c30dde0bb
+size 19014

dummy/wiki_lingua_german_de/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8bcfa7beb23d687c91be4ded92b92df8eddaccad78c88ecce7995206d95df5e
+size 17761

dummy/wiki_lingua_hindi_hi/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:208ebb495ce596e6c6f089c0e56c3dde89bb9fa1c33f8aa761c3c3f13388806e
+size 19685

dummy/wiki_lingua_indonesian_id/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88233d0425c7dfc79c1b8d391362aac9c9187be46510ce945f0dab7c5f9eab69
+size 17529

dummy/wiki_lingua_italian_it/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc45ca716a30d44aa48e471ca2323903f8a6c74c1f77cefdb1d76ed2f46415c7
+size 19783

dummy/wiki_lingua_japanese_ja/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ce04ea92ab7b9ac1ab1521df2e31c1eeb4cf62d72fda5a4d18c02797c919c07
+size 17113

dummy/wiki_lingua_korean_ko/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb813186e3e1470817745f88f16e801cd7cdeb529a7a4660b71e885139298a77
+size 18429

dummy/wiki_lingua_portuguese_pt/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9445b917df8e18396338b11c0d8593d6069166449ef7ef8bc51d3c06711449b
+size 19252

dummy/wiki_lingua_russian_ru/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13450e16cec76a371afde6da6ad11b2eb60a39f7eb99dd4b8d7165483b6fcbc3
+size 18047

dummy/wiki_lingua_spanish_es/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1052e008149de5507c6006ec31ff3bd9f94f0d3756cc2c3742d15c4eca9b417b
+size 18129

dummy/wiki_lingua_thai_th/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56e58e66d2e99394206f05f8e4cc6d5d488b3339c7c23cf59e6ce6f4cc346230
+size 17239

dummy/wiki_lingua_turkish_tr/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3aa9612fd7f32c5d741b6a260ea8eae4c898c66a738d44de6b9df7911aceca7c
+size 17698

dummy/wiki_lingua_vietnamese_vi/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25756c0fa718689d2e7f6948d58afc055431a1338c4c6e4de0d9b59f40269d5d
+size 21258

dummy/xsum/1.0.0/dummy_data.zip

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c5f62f61f9fdb8eed99b3368c890cfc148e950665e53957f575d4c2b65d9fc48
-size 2919

dummy/xsum/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1f81d5669e596bf21e4438bf909d134dc474c3e489bcff6e64434dff67b5427
+size 22590

gem.py

@@ -14,7 +14,6 @@
 # limitations under the License.
 """GEM: Generation Evaluation Metrics supporting datasets"""
 
-from __future__ import absolute_import, division, print_function
 
 import csv
 import json
@@ -23,13 +22,71 @@ import os
 import datasets
 
 
-# TODO: Add BibTeX citation
 _CITATION = """\
-@InProceedings{huggingface:dataset,
-title = {A great new dataset},
-authors={huggingface, Inc.
-},
-year={2020}
+@article{gem_benchmark,
+  author    = {Sebastian Gehrmann and
+               Tosin P. Adewumi and
+               Karmanya Aggarwal and
+               Pawan Sasanka Ammanamanchi and
+               Aremu Anuoluwapo and
+               Antoine Bosselut and
+               Khyathi Raghavi Chandu and
+               Miruna{-}Adriana Clinciu and
+               Dipanjan Das and
+               Kaustubh D. Dhole and
+               Wanyu Du and
+               Esin Durmus and
+               Ondrej Dusek and
+               Chris Emezue and
+               Varun Gangal and
+               Cristina Garbacea and
+               Tatsunori Hashimoto and
+               Yufang Hou and
+               Yacine Jernite and
+               Harsh Jhamtani and
+               Yangfeng Ji and
+               Shailza Jolly and
+               Dhruv Kumar and
+               Faisal Ladhak and
+               Aman Madaan and
+               Mounica Maddela and
+               Khyati Mahajan and
+               Saad Mahamood and
+               Bodhisattwa Prasad Majumder and
+               Pedro Henrique Martins and
+               Angelina McMillan{-}Major and
+               Simon Mille and
+               Emiel van Miltenburg and
+               Moin Nadeem and
+               Shashi Narayan and
+               Vitaly Nikolaev and
+               Rubungo Andre Niyongabo and
+               Salomey Osei and
+               Ankur P. Parikh and
+               Laura Perez{-}Beltrachini and
+               Niranjan Ramesh Rao and
+               Vikas Raunak and
+               Juan Diego Rodriguez and
+               Sashank Santhanam and
+               Joao Sedoc and
+               Thibault Sellam and
+               Samira Shaikh and
+               Anastasia Shimorina and
+               Marco Antonio Sobrevilla Cabezudo and
+               Hendrik Strobelt and
+               Nishant Subramani and
+               Wei Xu and
+               Diyi Yang and
+               Akhila Yerukola and
+               Jiawei Zhou},
+  title     = {The {GEM} Benchmark: Natural Language Generation, its Evaluation and
+               Metrics},
+  journal   = {CoRR},
+  volume    = {abs/2102.01672},
+  year      = {2021},
+  url       = {https://arxiv.org/abs/2102.01672},
+  archivePrefix = {arXiv},
+  eprint    = {2102.01672}
 }
 """
 
@@ -53,7 +110,30 @@ _LICENSE = "CC-BY-SA-4.0"
 _TASKS = {
     "summarization": {
         "mlsum": ["mlsum_de", "mlsum_es"],
-        "wiki_lingua": ["wiki_lingua_es_en", "wiki_lingua_ru_en", "wiki_lingua_tr_en", "wiki_lingua_vi_en"],
+        "wiki_lingua": [
+            "wiki_lingua_es_en_v0",
+            "wiki_lingua_ru_en_v0",
+            "wiki_lingua_tr_en_v0",
+            "wiki_lingua_vi_en_v0",
+            "wiki_lingua_arabic_ar",
+            "wiki_lingua_chinese_zh",
+            "wiki_lingua_czech_cs",
+            "wiki_lingua_dutch_nl",
+            "wiki_lingua_english_en",
+            "wiki_lingua_french_fr",
+            "wiki_lingua_german_de",
+            "wiki_lingua_hindi_hi",
+            "wiki_lingua_indonesian_id",
+            "wiki_lingua_italian_it",
+            "wiki_lingua_japanese_ja",
+            "wiki_lingua_korean_ko",
+            "wiki_lingua_portuguese_pt",
+            "wiki_lingua_russian_ru",
+            "wiki_lingua_spanish_es",
+            "wiki_lingua_thai_th",
+            "wiki_lingua_turkish_tr",
+            "wiki_lingua_vietnamese_vi",
+        ],
         "xsum": ["xsum"],
     },
     "struct2text": {
@@ -75,11 +155,13 @@ _TASKS = {
 _URLs = {
     "common_gen": {
         "data": "https://storage.googleapis.com/huggingface-nlp/datasets/common_gen/commongen_data.zip",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/common_gen.zip",
     },
     "cs_restaurants": {
         "train": "https://raw.githubusercontent.com/UFAL-DSG/cs_restaurant_dataset/master/train.json",
         "validation": "https://raw.githubusercontent.com/UFAL-DSG/cs_restaurant_dataset/master/devel.json",
         "test": "https://raw.githubusercontent.com/UFAL-DSG/cs_restaurant_dataset/master/test.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/cs_restaurants.zip",
     },
     "dart": {
         "train": "https://raw.githubusercontent.com/Yale-LILY/dart/master/data/v1.1.1/dart-v1.1.1-full-train.json",
@@ -90,68 +172,130 @@ _URLs = {
         "train": "https://github.com/tuetschek/e2e-cleaning/raw/master/cleaned-data/train-fixed.no-ol.csv",
         "validation": "https://github.com/tuetschek/e2e-cleaning/raw/master/cleaned-data/devel-fixed.no-ol.csv",
         "test": "https://github.com/tuetschek/e2e-cleaning/raw/master/cleaned-data/test-fixed.csv",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/e2e_nlg.zip",
     },
     "mlsum_de": {
         "train": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/de_train.zip",
         "validation": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/de_val.zip",
         "test": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/de_test.zip",
         "bad_ids": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_mlsum_bad_ids_fixed.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/mlsum_de.zip",
     },
     "mlsum_es": {
         "train": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/es_train.zip",
         "validation": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/es_val.zip",
         "test": "https://gitlab.lip6.fr/scialom/mlsum_data/-/raw/master/MLSUM/es_test.zip",
         "bad_ids": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_mlsum_bad_ids_fixed.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/mlsum_es.zip",
     },
     "schema_guided_dialog": {
-        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_sgd.json.zip",
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_sgd_context.zip",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/schema_guided_dialog.zip",
     },
     "totto": {
         "data": "https://storage.googleapis.com/totto/totto_data.zip",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/totto.zip",
     },
     "web_nlg_en": {
         "train": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_en_train.json",
         "validation": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_en_val.json",
         "test": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_en_test.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/web_nlg_en.zip",
     },
     "web_nlg_ru": {
         "train": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_ru_train.json",
         "validation": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_ru_val.json",
         "test": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_web_nlg/webnlg_ru_test.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/web_nlg_ru.zip",
     },
     "wiki_auto_asset_turk": {
-        "train": "https://github.com/chaojiang06/wiki-auto/raw/master/wiki-manual/train.tsv",
-        "validation": "https://github.com/chaojiang06/wiki-auto/raw/master/wiki-manual/dev.tsv",
+        "train": "https://github.com/chaojiang06/wiki-auto/raw/master/wiki-auto/GEM2021/full_with_split/train.tsv",
+        "validation": "https://github.com/chaojiang06/wiki-auto/raw/master/wiki-auto/GEM2021/full_with_split/valid.tsv",
+        "test_turk": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_turk_detokenized.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/wiki_auto_asset_turk_train_valid.zip",
     },
-    "wiki_lingua_es_en": {
+    "wiki_lingua_es_en_v0": {
         "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua.zip",
     },
-    "wiki_lingua_ru_en": {
+    "wiki_lingua_ru_en_v0": {
         "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua.zip",
     },
-    "wiki_lingua_tr_en": {
+    "wiki_lingua_tr_en_v0": {
         "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua.zip",
     },
-    "wiki_lingua_vi_en": {
+    "wiki_lingua_vi_en_v0": {
         "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua.zip",
     },
+    "wiki_lingua_arabic_ar": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/arabic.zip",
+    },
+    "wiki_lingua_chinese_zh": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/chinese.zip",
+    },
+    "wiki_lingua_czech_cs": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/czech.zip",
+    },
+    "wiki_lingua_dutch_nl": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/dutch.zip",
+    },
+    "wiki_lingua_english_en": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/english.zip",
+    },
+    "wiki_lingua_french_fr": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/french.zip",
+    },
+    "wiki_lingua_german_de": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/german.zip",
+    },
+    "wiki_lingua_hindi_hi": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/hindi.zip",
+    },
+    "wiki_lingua_indonesian_id": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/indonesian.zip",
+    },
+    "wiki_lingua_italian_it": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/italian.zip",
+    },
+    "wiki_lingua_japanese_ja": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/japanese.zip",
+    },
+    "wiki_lingua_korean_ko": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/korean.zip",
+    },
+    "wiki_lingua_portuguese_pt": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/portuguese.zip",
+    },
+    "wiki_lingua_russian_ru": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/russian.zip",
+    },
+    "wiki_lingua_spanish_es": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/spanish.zip",
+    },
+    "wiki_lingua_thai_th": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/thai.zip",
+    },
+    "wiki_lingua_turkish_tr": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/turkish.zip",
+    },
+    "wiki_lingua_vietnamese_vi": {
+        "data": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_wikilingua_full/vietnamese.zip",
+    },
     "xsum": {
         "data": "http://bollin.inf.ed.ac.uk/public/direct/XSUM-EMNLP18-Summary-Data-Original.tar.gz",
         "splits": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_xsum_confidence_0.8.json",
+        "challenge_set": "https://storage.googleapis.com/huggingface-nlp/datasets/gem/gem_challenge_sets/xsum.zip",
     },
 }
 
-# Add Turk and Asset files
+# Add Asset files
+_URLs["wiki_auto_asset_turk"][
+    "test_asset_orig"
+] = "https://github.com/facebookresearch/asset/raw/master/dataset/asset.test.orig"
 for i in range(10):
     _URLs["wiki_auto_asset_turk"][
         f"test_asset_{i}"
     ] = f"https://github.com/facebookresearch/asset/raw/master/dataset/asset.test.simp.{i}"
 
-for i in range(8):
-    _URLs["wiki_auto_asset_turk"][
-        f"test_turk_{i}"
-    ] = f"https://raw.githubusercontent.com/cocoxu/simplification/master/data/turkcorpus/GEM/test.8turkers.tok.turk.{i}"
-
 _SGD_ACTS = [
     "AFFIRM",
     "AFFIRM_INTENT",
@@ -196,7 +340,7 @@ class Gem(datasets.GeneratorBasedBuilder):
     BUILDER_CONFIGS = [
         datasets.BuilderConfig(
             name=conf,
-            version=datasets.Version("1.0.0"),
+            version=datasets.Version("1.1.0"),
             description=f"GEM benchmark: {task} task, {conf} subset",
         )
         for task, dset_confs in _TASKS.items()
@@ -211,6 +355,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "concept_set_id": datasets.Value("int32"),
                     "concepts": [datasets.Value("string")],
                     "target": datasets.Value("string"),  # single target for train
@@ -221,6 +366,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "dialog_act": datasets.Value("string"),
                     "dialog_act_delexicalized": datasets.Value("string"),
                     "target_delexicalized": datasets.Value("string"),
@@ -232,6 +378,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "dart_id": datasets.Value("int32"),
                     "tripleset": [[datasets.Value("string")]],  # list of triples
                     "subtree_was_extended": datasets.Value("bool"),
@@ -244,6 +391,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "meaning_representation": datasets.Value("string"),
                     "target": datasets.Value("string"),
                     "references": [datasets.Value("string")],
@@ -253,6 +401,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "text": datasets.Value("string"),
                     "topic": datasets.Value("string"),
                     "url": datasets.Value("string"),
@@ -266,6 +415,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "dialog_acts": [
                         {
                             "act": datasets.ClassLabel(names=_SGD_ACTS),
@@ -273,7 +423,9 @@ class Gem(datasets.GeneratorBasedBuilder):
                             "values": [datasets.Value("string")],
                         }
                     ],
+                    "context": [datasets.Value("string")],
                     "dialog_id": datasets.Value("string"),
+                    "service": datasets.Value("string"),
                     "turn_id": datasets.Value("int32"),
                     "prompt": datasets.Value("string"),
                     "target": datasets.Value("string"),
@@ -284,6 +436,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "totto_id": datasets.Value("int32"),
                     "table_page_title": datasets.Value("string"),
                     "table_webpage_url": datasets.Value("string"),
@@ -318,6 +471,7 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "input": [datasets.Value("string")],
                     "target": datasets.Value("string"),  # single target for train
                     "references": [datasets.Value("string")],
@@ -329,26 +483,41 @@ class Gem(datasets.GeneratorBasedBuilder):
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
-                    "source_id": datasets.Value("string"),
-                    "target_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "source": datasets.Value("string"),
                     "target": datasets.Value("string"),
                     "references": [datasets.Value("string")],
                 }
             )
         elif self.config.name.startswith("wiki_lingua"):
-            features = datasets.Features(
-                {
-                    "gem_id": datasets.Value("string"),
-                    "source": datasets.Value("string"),
-                    "target": datasets.Value("string"),
-                    "references": [datasets.Value("string")],
-                }
-            )
+            if "v0" in self.config.name:
+                features = datasets.Features(
+                    {
+                        "gem_id": datasets.Value("string"),
+                        "gem_parent_id": datasets.Value("string"),
+                        "source": datasets.Value("string"),
+                        "target": datasets.Value("string"),
+                        "references": [datasets.Value("string")],
+                    }
+                )
+            else:
+                ln = self.config.name.split("_")[-1]
+                features = datasets.Features(
+                    {
+                        "gem_id": datasets.Value("string"),
+                        "gem_parent_id": datasets.Value("string"),
+                        "source_aligned": datasets.Translation(languages=[ln, "en"]),
+                        "target_aligned": datasets.Translation(languages=[ln, "en"]),
+                        "source": datasets.Value("string"),
+                        "target": datasets.Value("string"),
+                        "references": [datasets.Value("string")],
+                    }
+                )
         elif self.config.name == "xsum":
             features = datasets.Features(
                 {
                     "gem_id": datasets.Value("string"),
+                    "gem_parent_id": datasets.Value("string"),
                     "xsum_id": datasets.Value("string"),
                     "document": datasets.Value("string"),
                     "target": datasets.Value("string"),
@@ -368,6 +537,11 @@ class Gem(datasets.GeneratorBasedBuilder):
         """Returns SplitGenerators."""
         dl_dir = dl_manager.download_and_extract(_URLs[self.config.name])
         if self.config.name == "common_gen":
+            challenge_sets = [
+                ("challenge_train_sample", "train_common_gen_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_common_gen_RandomSample500.json"),
+                ("challenge_test_scramble", "test_common_gen_ScrambleInputStructure500.json"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=datasets.Split.TRAIN,
@@ -390,11 +564,34 @@ class Gem(datasets.GeneratorBasedBuilder):
                         "split": "test",
                     },
                 ),
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name == "cs_restaurants":
+            challenge_sets = [
+                ("challenge_train_sample", "train_cs_restaurants_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_cs_restaurants_RandomSample500.json"),
+                ("challenge_test_scramble", "test_cs_restaurants_ScrambleInputStructure500.json"),
+            ]
             return [
                 datasets.SplitGenerator(name=spl, gen_kwargs={"filepath": dl_dir[spl], "split": spl})
                 for spl in ["train", "validation", "test"]
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name == "dart":
             return [
@@ -402,12 +599,31 @@ class Gem(datasets.GeneratorBasedBuilder):
                 for spl in ["train", "validation", "test"]
             ]
         elif self.config.name == "e2e_nlg":
+            challenge_sets = [
+                ("challenge_train_sample", "train_e2e_nlg_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_e2e_nlg_RandomSample500.json"),
+                ("challenge_test_scramble", "test_e2e_nlg_ScrambleInputStructure500.json"),
+            ]
             return [
                 datasets.SplitGenerator(name=spl, gen_kwargs={"filepath": dl_dir[spl], "split": spl})
                 for spl in ["train", "validation", "test"]
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name.startswith("mlsum"):
             lang = self.config.name.split("_")[1]
+            challenge_sets = [
+                ("challenge_train_sample", f"train_mlsum_{lang}_RandomSample500.json"),
+                ("challenge_validation_sample", f"validation_mlsum_{lang}_RandomSample500.json"),
+                ("challenge_test_covid", f"{lang}_test_covid19_cleaned.jsonl"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=datasets.Split.TRAIN,
@@ -436,15 +652,53 @@ class Gem(datasets.GeneratorBasedBuilder):
                         "filepaths": dl_dir["bad_ids"],
                     },
                 ),
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name == "schema_guided_dialog":
+            challenge_sets = [
+                ("challenge_train_sample", "train_schema_guided_dialog_RandomSample500_reformatted.json"),
+                ("challenge_validation_sample", "validation_schema_guided_dialog_RandomSample500_reformatted.json"),
+                ("challenge_test_backtranslation", "test_schema_guided_dialog_BackTranslation500_reformatted.json"),
+                (
+                    "challenge_test_bfp02",
+                    "test_schema_guided_dialog_ButterFingersPerturbation_p=0.02_500_reformatted.json",
+                ),
+                (
+                    "challenge_test_bfp05",
+                    "test_schema_guided_dialog_ButterFingersPerturbation_p=0.05_500_reformatted.json",
+                ),
+                ("challenge_test_nopunc", "test_schema_guided_dialog_WithoutPunctuation500_reformatted.json"),
+                ("challenge_test_scramble", "test_schema_guided_dialog_ScrambleInputStructure500_reformatted.json"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=spl, gen_kwargs={"filepath": os.path.join(dl_dir["data"], "gem_sgd.json"), "split": spl}
                 )
                 for spl in ["train", "validation", "test"]
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name == "totto":
+            challenge_sets = [
+                ("challenge_train_sample", "train_totto_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_totto_RandomSample500.json"),
+                ("challenge_test_scramble", "test_totto_ScrambleInputStructure500.json"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=datasets.Split.TRAIN,
@@ -467,13 +721,63 @@ class Gem(datasets.GeneratorBasedBuilder):
                         "split": "test",
                     },
                 ),
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name.startswith("web_nlg"):
+            ln = self.config.name.split("_")[-1]
+            challenge_sets = [
+                ("challenge_train_sample", f"train_web_nlg_{ln}_RandomSample500.json"),
+                ("challenge_validation_sample", f"validation_web_nlg_{ln}_RandomSample500.json"),
+                ("challenge_test_scramble", f"test_web_nlg_{ln}_ScrambleInputStructure500.json"),
+            ]
+            if ln == "en":
+                challenge_sets += [("challenge_test_numbers", f"test_web_nlg_{ln}_replace_numbers_500.json")]
             return [
                 datasets.SplitGenerator(name=spl, gen_kwargs={"filepath": dl_dir[spl], "split": spl})
                 for spl in ["train", "validation", "test"]
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], self.config.name, filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
         elif self.config.name == "wiki_auto_asset_turk":
+            challenge_sets = [
+                ("challenge_train_sample", "train_wiki_auto_asset_turk_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_wiki_auto_asset_turk_RandomSample500.json"),
+                ("challenge_test_asset_backtranslation", "test_asset_wiki_auto_asset_turk_BackTranslation.json"),
+                (
+                    "challenge_test_asset_bfp02",
+                    "test_asset_wiki_auto_asset_turk_ButterFingersPerturbation_p=0.02.json",
+                ),
+                (
+                    "challenge_test_asset_bfp05",
+                    "test_asset_wiki_auto_asset_turk_ButterFingersPerturbation_p=0.05.json",
+                ),
+                ("challenge_test_asset_nopunc", "test_asset_wiki_auto_asset_turk_WithoutPunctuation.json"),
+                ("challenge_test_turk_backtranslation", "detok_test_turk_wiki_auto_asset_turk_BackTranslation.json"),
+                (
+                    "challenge_test_turk_bfp02",
+                    "detok_test_turk_wiki_auto_asset_turk_ButterFingersPerturbation_p=0.02.json",
+                ),
+                (
+                    "challenge_test_turk_bfp05",
+                    "detok_test_turk_wiki_auto_asset_turk_ButterFingersPerturbation_p=0.05.json",
+                ),
+                ("challenge_test_turk_nopunc", "detok_test_turk_wiki_auto_asset_turk_WithoutPunctuation.json"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=datasets.Split.TRAIN,
@@ -493,46 +797,94 @@ class Gem(datasets.GeneratorBasedBuilder):
                     name="test_asset",
                     gen_kwargs={
                         "filepath": "",
-                        "split": "test",
-                        "filepaths": [dl_dir[f"test_asset_{i}"] for i in range(10)],
+                        "split": "test_asset",
+                        "filepaths": [dl_dir["test_asset_orig"]] + [dl_dir[f"test_asset_{i}"] for i in range(10)],
                     },
                 ),
                 datasets.SplitGenerator(
                     name="test_turk",
                     gen_kwargs={
-                        "filepath": "",
-                        "split": "test",
-                        "filepaths": [dl_dir[f"test_turk_{i}"] for i in range(8)],
+                        "filepath": dl_dir["test_turk"],
+                        "split": "test_turk",
                     },
                 ),
-            ]
-        elif self.config.name.startswith("wiki_lingua"):
-            lang = self.config.name.split("_")[-2]
-            base_dir = os.path.join(dl_dir["data"], "GEM_data_crosslingual", f"{lang}_en")
-            return [
+            ] + [
                 datasets.SplitGenerator(
-                    name=datasets.Split.TRAIN,
+                    name=challenge_split,
                     gen_kwargs={
-                        "filepath": base_dir,
-                        "split": "train",
+                        "filepath": os.path.join(dl_dir["challenge_set"], "wiki_auto_asset_turk", filename),
+                        "split": challenge_split,
                     },
-                ),
-                datasets.SplitGenerator(
-                    name=datasets.Split.VALIDATION,
-                    gen_kwargs={
-                        "filepath": base_dir,
-                        "split": "val",
-                    },
-                ),
-                datasets.SplitGenerator(
-                    name=datasets.Split.TEST,
-                    gen_kwargs={
-                        "filepath": base_dir,
-                        "split": "test",
-                    },
-                ),
+                )
+                for challenge_split, filename in challenge_sets
             ]
+        elif self.config.name.startswith("wiki_lingua"):
+            if "v0" in self.config.name:
+                lang = self.config.name.split("_")[-3]
+                base_dir = os.path.join(dl_dir["data"], "GEM_data_crosslingual", f"{lang}_en")
+                return [
+                    datasets.SplitGenerator(
+                        name=datasets.Split.TRAIN,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "train",
+                        },
+                    ),
+                    datasets.SplitGenerator(
+                        name=datasets.Split.VALIDATION,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "val",
+                        },
+                    ),
+                    datasets.SplitGenerator(
+                        name=datasets.Split.TEST,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "test",
+                        },
+                    ),
+                ]
+            else:
+                lang_name = self.config.name.split("_")[-2]
+                lang = self.config.name.split("_")[-1]
+                base_dir = os.path.join(dl_dir["data"], lang_name)
+                return [
+                    datasets.SplitGenerator(
+                        name=datasets.Split.TRAIN,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "train",
+                            "lang": lang,
+                        },
+                    ),
+                    datasets.SplitGenerator(
+                        name=datasets.Split.VALIDATION,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "val",
+                            "lang": lang,
+                        },
+                    ),
+                    datasets.SplitGenerator(
+                        name=datasets.Split.TEST,
+                        gen_kwargs={
+                            "filepath": base_dir,
+                            "split": "test",
+                            "lang": lang,
+                        },
+                    ),
+                ]
         elif self.config.name == "xsum":
+            challenge_sets = [
+                ("challenge_train_sample", "train_xsum_RandomSample500.json"),
+                ("challenge_validation_sample", "validation_xsum_RandomSample500.json"),
+                ("challenge_test_backtranslation", "test_xsum_BackTranslation500.json"),
+                ("challenge_test_bfp_02", "test_xsum_ButterFingersPerturbation_p=0.02_500.json"),
+                ("challenge_test_bfp_05", "test_xsum_ButterFingersPerturbation_p=0.05_500.json"),
+                ("challenge_test_nopunc", "test_xsum_WithoutPunctuation500.json"),
+                ("challenge_test_covid", f"en_test_covid19.jsonl"),
+            ]
             return [
                 datasets.SplitGenerator(
                     name=datasets.Split.TRAIN,
@@ -558,50 +910,86 @@ class Gem(datasets.GeneratorBasedBuilder):
                         "filepaths": os.path.join(dl_dir["data"], "bbc-summary-data"),
                     },
                 ),
+            ] + [
+                datasets.SplitGenerator(
+                    name=challenge_split,
+                    gen_kwargs={
+                        "filepath": os.path.join(dl_dir["challenge_set"], "xsum", filename),
+                        "split": challenge_split,
+                    },
+                )
+                for challenge_split, filename in challenge_sets
             ]
 
     def _generate_examples(self, filepath, split, filepaths=None, lang=None):
         """ Yields examples. """
         if self.config.name == "common_gen":
-            with open(filepath, encoding="utf-8") as f:
-                id_ = -1
-                i = -1
-                for row in f:
-                    row = row.replace(", }", "}")  # Fix possible JSON format error
-                    data = json.loads(row)
-                    concepts = [word for word in data["concept_set"].split("#")]
-                    if split == "train":
-                        i += 1
-                        for scene in data["scene"]:
+            if split.startswith("challenge"):
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                with open(filepath, encoding="utf-8") as f:
+                    id_ = -1
+                    i = -1
+                    for row in f:
+                        row = row.replace(", }", "}")  # Fix possible JSON format error
+                        data = json.loads(row)
+                        concepts = [word for word in data["concept_set"].split("#")]
+                        if split == "train":
+                            i += 1
+                            for scene in data["scene"]:
+                                id_ += 1
+                                yield id_, {
+                                    "gem_id": f"{self.config.name}-{split}-{id_}",
+                                    "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                    "concept_set_id": i,
+                                    "concepts": concepts,
+                                    "target": scene,
+                                    "references": [],
+                                }
+                        else:
                             id_ += 1
                             yield id_, {
                                 "gem_id": f"{self.config.name}-{split}-{id_}",
-                                "concept_set_id": i,
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                "concept_set_id": id_,
                                 "concepts": concepts,
-                                "target": scene,
-                                "references": [],
+                                "target": "" if split == "test" else data["scene"][0],
+                                "references": [] if split == "test" else data["scene"],
                             }
-                    else:
-                        id_ += 1
+        elif self.config.name == "cs_restaurants":
+            if split.startswith("challenge"):
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                with open(filepath, encoding="utf8") as f:
+                    data = json.load(f)
+                    for id_, instance in enumerate(data):
                         yield id_, {
                             "gem_id": f"{self.config.name}-{split}-{id_}",
-                            "concept_set_id": id_,
-                            "concepts": concepts,
-                            "target": "" if split == "test" else data["scene"][0],
-                            "references": [] if split == "test" else data["scene"],
+                            "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                            "dialog_act": instance["da"],
+                            "dialog_act_delexicalized": instance["delex_da"],
+                            "target": instance["text"],
+                            "target_delexicalized": instance["delex_text"],
+                            "references": [] if split == "train" else [instance["text"]],
                         }
-        elif self.config.name == "cs_restaurants":
-            with open(filepath, encoding="utf8") as f:
-                data = json.load(f)
-                for id_, instance in enumerate(data):
-                    yield id_, {
-                        "gem_id": f"{self.config.name}-{split}-{id_}",
-                        "dialog_act": instance["da"],
-                        "dialog_act_delexicalized": instance["delex_da"],
-                        "target": instance["text"],
-                        "target_delexicalized": instance["delex_text"],
-                        "references": [] if split == "train" else [instance["text"]],
-                    }
         elif self.config.name == "dart":
             with open(filepath, encoding="utf-8") as f:
                 data = json.loads(f.read())
@@ -614,6 +1002,7 @@ class Gem(datasets.GeneratorBasedBuilder):
                             id_ += 1
                             yield id_, {
                                 "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
                                 "dart_id": i,
                                 "tripleset": example["tripleset"],
                                 "subtree_was_extended": example.get("subtree_was_extended", None),  # some are missing
@@ -625,6 +1014,7 @@ class Gem(datasets.GeneratorBasedBuilder):
                         id_ += 1
                         yield id_, {
                             "gem_id": f"{self.config.name}-{split}-{id_}",
+                            "gem_parent_id": f"{self.config.name}-{split}-{id_}",
                             "dart_id": id_,
                             "tripleset": example["tripleset"],
                             "subtree_was_extended": example.get("subtree_was_extended", None),  # some are missing
@@ -633,69 +1023,145 @@ class Gem(datasets.GeneratorBasedBuilder):
                             "references": [annotation["text"] for annotation in example["annotations"]],
                         }
         elif self.config.name == "e2e_nlg":
-            with open(filepath, encoding="utf-8") as f:
-                reader = csv.DictReader(f)
-                for id_, example in enumerate(reader):
-                    yield id_, {
-                        "gem_id": f"{self.config.name}-{split}-{id_}",
-                        "meaning_representation": example["mr"],
-                        "target": example["ref"],
-                        "references": [] if split == "train" else [example["ref"]],
-                    }
-        elif self.config.name.startswith("mlsum"):
-            bad_ids_dct = json.load(open(filepaths, encoding="utf-8"))
-            bad_ids = dict((bad_url, True) for _, bad_url in bad_ids_dct[f"{lang}-{split}"])
-            with open(filepath, encoding="utf-8") as f:
-                id_ = -1
-                for line in f:
-                    data = json.loads(line)
-                    if data["url"] in bad_ids:  # TODO : check | i or i-1?
+            if split.startswith("challenge"):
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
                         continue
-                    else:
-                        id_ += 1
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                with open(filepath, encoding="utf-8") as f:
+                    reader = csv.DictReader(f)
+                    for id_, example in enumerate(reader):
                         yield id_, {
                             "gem_id": f"{self.config.name}-{split}-{id_}",
-                            "text": data["text"],
-                            "target": data["summary"],
-                            "references": [] if split == "train" else [data["summary"]],
-                            "topic": data["topic"],
-                            "url": data["url"],
-                            "title": data["title"],
-                            "date": data["date"],
+                            "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                            "meaning_representation": example["mr"],
+                            "target": example["ref"],
+                            "references": [] if split == "train" else [example["ref"]],
                         }
+        elif self.config.name.startswith("mlsum"):
+            if split in ["train", "validation", "test", "challenge_test_covid"]:
+                if split == "challenge_test_covid":
+                    bad_ids = {}
+                else:
+                    bad_ids_dct = json.load(open(filepaths, encoding="utf-8"))
+                    bad_ids = dict((bad_url, True) for _, bad_url in bad_ids_dct[f"{lang}-{split}"])
+                with open(filepath, encoding="utf-8") as f:
+                    id_ = -1
+                    for line in f:
+                        data = json.loads(line)
+                        if data["url"] in bad_ids:
+                            continue
+                        else:
+                            id_ += 1
+                            yield id_, {
+                                "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                "text": data["text"],
+                                "target": data["summary"],
+                                "references": [] if split == "train" else [data["summary"]],
+                                "topic": data["topic"],
+                                "url": data["url"],
+                                "title": data["title"],
+                                "date": data["date"],
+                            }
+            else:
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
         elif self.config.name == "schema_guided_dialog":
-            examples = json.load(open(filepath, encoding="utf-8"))[split]
-            for id_, example in enumerate(examples):
-                yield id_, {
-                    "gem_id": f"{self.config.name}-{split}-{id_}",
-                    "dialog_acts": [
-                        {
-                            "act": act_id,
-                            "slot": slot,
-                            "values": values,
-                        }
-                        for act_id, slot, values in example["da"]
-                    ],
-                    "dialog_id": example["dialog_id"],
-                    "turn_id": example["turn_ix"],
-                    "prompt": example["prompt"],
-                    "target": example["target"],
-                    "references": [] if split == "train" else [example["target"]],
-                }
+            if "challenge" in split:
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                examples = json.load(open(filepath, encoding="utf-8"))[split]
+                for id_, example in enumerate(examples):
+                    yield id_, {
+                        "gem_id": f"{self.config.name}-{split}-{id_}",
+                        "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                        "dialog_acts": [
+                            {
+                                "act": act_id,
+                                "slot": slot,
+                                "values": values,
+                            }
+                            for act_id, slot, values in example["da"]
+                        ],
+                        "context": example["context"],
+                        "dialog_id": example["dialog_id"],
+                        "service": example["service"],
+                        "turn_id": example["turn_ix"],
+                        "prompt": example["prompt"],
+                        "target": example["target"],
+                        "references": [] if split == "train" else [example["target"]],
+                    }
         elif self.config.name == "totto":
-            with open(filepath, "r", encoding="utf-8") as json_file:
-                json_list = list(json_file)
-            id_ = -1
-            i = -1
-            for json_str in json_list:
-                result = json.loads(json_str)
-                if split == "train":
-                    i += 1
-                    for sentence in result["sentence_annotations"]:
+            if "challenge" in split:
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                with open(filepath, "r", encoding="utf-8") as json_file:
+                    json_list = list(json_file)
+                id_ = -1
+                i = -1
+                for json_str in json_list:
+                    result = json.loads(json_str)
+                    if split == "train":
+                        i += 1
+                        for sentence in result["sentence_annotations"]:
+                            id_ += 1
+                            response = {
+                                "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                "totto_id": i,
+                                "table_page_title": result["table_page_title"],
+                                "table_webpage_url": result["table_webpage_url"],
+                                "table_section_title": result["table_section_title"],
+                                "table_section_text": result["table_section_text"],
+                                "table": result["table"],
+                                "highlighted_cells": result["highlighted_cells"],
+                                "example_id": str(result["example_id"]),
+                                "overlap_subset": "none",
+                                "sentence_annotations": [sentence],
+                                "references": [],
+                                "target": sentence["final_sentence"],
+                            }
+                            yield id_, response
+                    else:
                         id_ += 1
                         response = {
                             "gem_id": f"{self.config.name}-{split}-{id_}",
-                            "totto_id": i,
+                            "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                            "totto_id": id_,
                             "table_page_title": result["table_page_title"],
                             "table_webpage_url": result["table_webpage_url"],
                             "table_section_title": result["table_section_title"],
@@ -703,106 +1169,167 @@ class Gem(datasets.GeneratorBasedBuilder):
                             "table": result["table"],
                             "highlighted_cells": result["highlighted_cells"],
                             "example_id": str(result["example_id"]),
-                            "overlap_subset": "none",
-                            "sentence_annotations": [sentence],
-                            "references": [],
-                            "target": sentence["final_sentence"],
+                            "overlap_subset": str(result["overlap_subset"]),
                         }
+                        response["sentence_annotations"] = [] if split == "test" else result["sentence_annotations"]
+                        response["references"] = [
+                            sentence["final_sentence"] for sentence in response["sentence_annotations"]
+                        ]
+                        response["target"] = response["references"][0] if len(response["references"]) > 0 else ""
                         yield id_, response
-                else:
-                    id_ += 1
-                    response = {
-                        "gem_id": f"{self.config.name}-{split}-{id_}",
-                        "totto_id": id_,
-                        "table_page_title": result["table_page_title"],
-                        "table_webpage_url": result["table_webpage_url"],
-                        "table_section_title": result["table_section_title"],
-                        "table_section_text": result["table_section_text"],
-                        "table": result["table"],
-                        "highlighted_cells": result["highlighted_cells"],
-                        "example_id": str(result["example_id"]),
-                        "overlap_subset": str(result["overlap_subset"]),
-                    }
-                    response["sentence_annotations"] = [] if split == "test" else result["sentence_annotations"]
-                    response["references"] = [
-                        sentence["final_sentence"] for sentence in response["sentence_annotations"]
-                    ]
-                    response["target"] = response["references"][0] if len(response["references"]) > 0 else ""
-                    yield id_, response
         elif self.config.name.startswith("web_nlg"):
-            with open(filepath, encoding="utf-8") as f:
-                examples = json.load(f)
-                id_ = -1
-                for example in examples["values"]:
-                    if split == "train":
-                        for target in example["target"]:
+            if "challenge" in split:
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    if len(exple) == 0:
+                        continue
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    yield id_, exple
+            else:
+                with open(filepath, encoding="utf-8") as f:
+                    examples = json.load(f)
+                    id_ = -1
+                    for example in examples["values"]:
+                        if split == "train":
+                            for target in example["target"]:
+                                id_ += 1
+                                yield id_, {
+                                    "gem_id": f"{self.config.name}-{split}-{id_}",
+                                    "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                    "input": example["input"],
+                                    "target": target,
+                                    "references": [] if split == "train" else example["target"],
+                                    "category": example["category"],
+                                    "webnlg_id": example["webnlg-id"],
+                                }
+                        else:
                             id_ += 1
                             yield id_, {
                                 "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
                                 "input": example["input"],
-                                "target": target,
-                                "references": [] if split == "train" else example["target"],
+                                "target": example["target"][0] if len(example["target"]) > 0 else "",
+                                "references": example["target"],
                                 "category": example["category"],
                                 "webnlg_id": example["webnlg-id"],
                             }
-                    else:
-                        id_ += 1
-                        yield id_, {
-                            "gem_id": f"{self.config.name}-{split}-{id_}",
-                            "input": example["input"],
-                            "target": example["target"][0] if len(example["target"]) > 0 else "",
-                            "references": example["target"],
-                            "category": example["category"],
-                            "webnlg_id": example["webnlg-id"],
-                        }
         elif self.config.name == "wiki_auto_asset_turk":
             if split in ["train", "validation"]:
                 keys = [
-                    "target_id",
-                    "source_id",
-                    "target",
                     "source",
+                    "target",
                 ]
                 with open(filepath, encoding="utf-8") as f:
                     for id_, line in enumerate(f):
                         values = line.strip().split("\t")
-                        assert len(values) == 5, f"Not enough fields in ---- {line} --- {values}"
-                        example = dict([(k, val) for k, val in zip(keys, values[1:])])
+                        assert len(values) == 2, f"Not enough fields in ---- {line} --- {values}"
+                        example = dict([(k, val) for k, val in zip(keys, values)])
                         example["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                        example["gem_parent_id"] = example["gem_id"]
                         example["references"] = [] if split == "train" else [example["target"]]
                         yield id_, example
-            elif split.startswith("test"):
+            elif split == "test_turk":
+                examples = json.load(open(filepath, encoding="utf-8"))
+                for id_, example in enumerate(examples):
+                    example["gem_parent_id"] = example["gem_id"]
+                    for k in ["source_id", "target_id"]:
+                        if k in example:
+                            del example[k]
+                    yield id_, example
+            elif split == "test_asset":
                 files = [open(f_name, encoding="utf-8") for f_name in filepaths]
                 for id_, lines in enumerate(zip(*files)):
                     yield id_, {
                         "gem_id": f"{self.config.name}-{split}-{id_}",
-                        "source_id": "",
-                        "target_id": "",
+                        "gem_parent_id": f"{self.config.name}-{split}-{id_}",
                         "target": lines[1].strip(),
                         "source": lines[0].strip(),
                         "references": [line.strip() for line in lines[1:]],
                     }
+            else:
+                exples = json.load(open(filepath, encoding="utf-8"))
+                if isinstance(exples, dict):
+                    assert len(exples) == 1, "multiple entries found"
+                    exples = list(exples.values())[0]
+                for id_, exple in enumerate(exples):
+                    exple["gem_parent_id"] = exple["gem_id"]
+                    exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                    for k in ["source_id", "target_id"]:
+                        if k in exple:
+                            del exple[k]
+                    yield id_, exple
         elif self.config.name.startswith("wiki_lingua"):
-            with open(os.path.join(filepath, f"{split}.src"), encoding="utf-8") as f_in:
-                with open(os.path.join(filepath, f"{split}.tgt"), encoding="utf-8") as f_out:
-                    for id_, (src, tgt) in enumerate(zip(f_in, f_out)):
+            if "v0" in self.config.name:
+                with open(os.path.join(filepath, f"{split}.src"), encoding="utf-8") as f_in:
+                    with open(os.path.join(filepath, f"{split}.tgt"), encoding="utf-8") as f_out:
+                        for id_, (src, tgt) in enumerate(zip(f_in, f_out)):
+                            yield id_, {
+                                "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                "source": src.strip(),
+                                "target": tgt.strip(),
+                                "references": [] if split == "train" else [tgt.strip()],
+                            }
+            else:
+                with open(os.path.join(filepath, f"{split}.src.{lang}"), encoding="utf-8") as f_in_ln:
+                    with open(os.path.join(filepath, f"{split}.src.en"), encoding="utf-8") as f_in_en:
+                        with open(os.path.join(filepath, f"{split}.tgt.{lang}"), encoding="utf-8") as f_out_ln:
+                            with open(os.path.join(filepath, f"{split}.tgt.en"), encoding="utf-8") as f_out_en:
+                                for id_, (src_ln, src_en, tgt_ln, tgt_en) in enumerate(
+                                    zip(f_in_ln, f_in_en, f_out_ln, f_out_en)
+                                ):
+                                    yield id_, {
+                                        "gem_id": f"{self.config.name}-{split}-{id_}",
+                                        "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                        "source_aligned": {lang: src_ln.strip(), "en": src_en.strip()},
+                                        "target_aligned": {lang: tgt_ln.strip(), "en": tgt_en.strip()},
+                                        "source": src_ln.strip(),
+                                        "target": tgt_en.strip(),
+                                        "references": [] if split == "train" else [tgt_en.strip()],
+                                    }
+        elif self.config.name == "xsum":
+            if "challenge" in split:
+                if "covid" in split:
+                    with open(filepath, encoding="utf-8") as f:
+                        id_ = -1
+                        for line in f:
+                            data = json.loads(line)
+                            id_ += 1
+                            yield id_, {
+                                "gem_id": f"{self.config.name}-{split}-{id_}",
+                                "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                                "xsum_id": data["url"],
+                                "document": data["text"],
+                                "target": data["summary"],
+                                "references": [] if split == "train" else [data["summary"]],
+                            }
+                else:
+                    exples = json.load(open(filepath, encoding="utf-8"))
+                    if isinstance(exples, dict):
+                        assert len(exples) == 1, "multiple entries found"
+                        exples = list(exples.values())[0]
+                    for id_, exple in enumerate(exples):
+                        exple["gem_parent_id"] = exple["gem_id"]
+                        exple["gem_id"] = f"{self.config.name}-{split}-{id_}"
+                        yield id_, exple
+            else:
+                with open(filepath, "r", encoding="utf-8") as f:
+                    split_ids = json.load(f)
+                for id_, i in enumerate(split_ids[split]):
+                    with open(os.path.join(filepaths, i + ".summary"), "r", encoding="utf-8") as f:
+                        text = "".join(
+                            [line for line in f.readlines() if line not in _XSUM_REMOVE_LINES and line.strip()]
+                        )
+                        segs = text.split("[SN]")
                         yield id_, {
                             "gem_id": f"{self.config.name}-{split}-{id_}",
-                            "source": src.strip(),
-                            "target": tgt.strip(),
-                            "references": [] if split == "train" else [tgt.strip()],
+                            "gem_parent_id": f"{self.config.name}-{split}-{id_}",
+                            "xsum_id": i,
+                            "document": segs[8].strip(),
+                            "target": segs[6].strip(),
+                            "references": [] if split == "train" else [segs[6].strip()],
                         }
-        elif self.config.name == "xsum":
-            with open(filepath, "r", encoding="utf-8") as f:
-                split_ids = json.load(f)
-            for id_, i in enumerate(split_ids[split]):
-                with open(os.path.join(filepaths, i + ".summary"), "r", encoding="utf-8") as f:
-                    text = "".join([line for line in f.readlines() if line not in _XSUM_REMOVE_LINES and line.strip()])
-                    segs = text.split("[SN]")
-                    yield id_, {
-                        "gem_id": f"{self.config.name}-{split}-{id_}",
-                        "xsum_id": i,
-                        "document": segs[8].strip(),
-                        "target": segs[6].strip(),
-                        "references": [] if split == "train" else [segs[6].strip()],
-                    }