add script, readme, dataset info and dummy data

.gitignore

@@ -0,0 +1,2 @@
+desktop.ini
+*.lock

README.md

@@ -0,0 +1,273 @@
+---
+pretty_name: TyDi QA
+annotations_creators:
+- crowdsourced
+language_creators:
+- crowdsourced
+languages:
+- en
+- ar
+- bn
+- fi
+- id
+- ja
+- sw
+- ko
+- ru
+- te
+- th
+licenses:
+- apache-2.0
+multilinguality:
+- multilingual
+size_categories:
+- unknown
+source_datasets:
+- extended|wikipedia
+task_categories:
+- question-answering
+task_ids:
+- extractive-qa
+paperswithcode_id: tydi-qa
+---
+
+# Dataset Card for "tydiqa"
+
+## Table of Contents
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+
+- **Homepage:** [https://github.com/google-research-datasets/tydiqa](https://github.com/google-research-datasets/tydiqa)
+- **Repository:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+- **Paper:** [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)
+- **Size of downloaded dataset files:** 3726.74 MB
+- **Size of the generated dataset:** 5812.92 MB
+- **Total amount of disk used:** 9539.67 MB
+
+### Dataset Summary
+
+TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.
+The languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language
+expresses -- such that we expect models performing well on this set to generalize across a large number of the languages
+in the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic
+information-seeking task and avoid priming effects, questions are written by people who want to know the answer, but
+don’t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without
+the use of translation (unlike MLQA and XQuAD).
+
+We also include "translate-train" and "translate-test" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The "translate-train" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.
+
+### Supported Tasks and Leaderboards
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Languages
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+## Dataset Structure
+
+### Data Instances
+
+#### primary_task
+
+- **Size of downloaded dataset files:** 1863.37 MB
+- **Size of the generated dataset:** 5757.59 MB
+- **Total amount of disk used:** 7620.96 MB
+
+An example of 'validation' looks as follows.
+```
+This example was too long and was cropped:
+
+{
+    "annotations": {
+        "minimal_answers_end_byte": [-1, -1, -1],
+        "minimal_answers_start_byte": [-1, -1, -1],
+        "passage_answer_candidate_index": [-1, -1, -1],
+        "yes_no_answer": ["NONE", "NONE", "NONE"]
+    },
+    "document_plaintext": "\"\\nรองศาสตราจารย์[1] หม่อมราชวงศ์สุขุมพันธุ์ บริพัตร  (22 กันยายน 2495 -) ผู้ว่าราชการกรุงเทพมหานครคนที่ 15 อดีตรองหัวหน้าพรรคปร...",
+    "document_title": "หม่อมราชวงศ์สุขุมพันธุ์ บริพัตร",
+    "document_url": "\"https://th.wikipedia.org/wiki/%E0%B8%AB%E0%B8%A1%E0%B9%88%E0%B8%AD%E0%B8%A1%E0%B8%A3%E0%B8%B2%E0%B8%8A%E0%B8%A7%E0%B8%87%E0%B8%...",
+    "language": "thai",
+    "passage_answer_candidates": "{\"plaintext_end_byte\": [494, 1779, 2931, 3904, 4506, 5588, 6383, 7122, 8224, 9375, 10473, 12563, 15134, 17765, 19863, 21902, 229...",
+    "question_text": "\"หม่อมราชวงศ์สุขุมพันธุ์ บริพัตร เรียนจบจากที่ไหน ?\"..."
+}
+```
+
+#### secondary_task
+
+- **Size of downloaded dataset files:** 1863.37 MB
+- **Size of the generated dataset:** 55.34 MB
+- **Total amount of disk used:** 1918.71 MB
+
+An example of 'validation' looks as follows.
+```
+This example was too long and was cropped:
+
+{
+    "answers": {
+        "answer_start": [394],
+        "text": ["بطولتين"]
+    },
+    "context": "\"أقيمت البطولة 21 مرة، شارك في النهائيات 78 دولة، وعدد الفرق التي فازت بالبطولة حتى الآن 8 فرق، ويعد المنتخب البرازيلي الأكثر تت...",
+    "id": "arabic-2387335860751143628-1",
+    "question": "\"كم عدد مرات فوز الأوروغواي ببطولة كاس العالم لكرو القدم؟\"...",
+    "title": "قائمة نهائيات كأس العالم"
+}
+```
+
+### Data Fields
+
+The data fields are the same among all splits.
+
+#### primary_task
+- `passage_answer_candidates`: a dictionary feature containing:
+  - `plaintext_start_byte`: a `int32` feature.
+  - `plaintext_end_byte`: a `int32` feature.
+- `question_text`: a `string` feature.
+- `document_title`: a `string` feature.
+- `language`: a `string` feature.
+- `annotations`: a dictionary feature containing:
+  - `passage_answer_candidate_index`: a `int32` feature.
+  - `minimal_answers_start_byte`: a `int32` feature.
+  - `minimal_answers_end_byte`: a `int32` feature.
+  - `yes_no_answer`: a `string` feature.
+- `document_plaintext`: a `string` feature.
+- `document_url`: a `string` feature.
+
+#### secondary_task
+- `id`: a `string` feature.
+- `title`: a `string` feature.
+- `context`: a `string` feature.
+- `question`: a `string` feature.
+- `answers`: a dictionary feature containing:
+  - `text`: a `string` feature.
+  - `answer_start`: a `int32` feature.
+
+### Data Splits
+
+| name           |  train | validation |
+| -------------- | -----: | ---------: |
+| primary_task   | 166916 |      18670 |
+| secondary_task |  49881 |       5077 |
+
+## Dataset Creation
+
+### Curation Rationale
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+#### Who are the source language producers?
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Annotations
+
+#### Annotation process
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+#### Who are the annotators?
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Personal and Sensitive Information
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+## Considerations for Using the Data
+
+### Social Impact of Dataset
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Discussion of Biases
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Other Known Limitations
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+## Additional Information
+
+### Dataset Curators
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Licensing Information
+
+[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
+
+### Citation Information
+
+```
+@article{tydiqa,
+title   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},
+author  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}
+year    = {2020},
+journal = {Transactions of the Association for Computational Linguistics}
+}
+
+
+
+
+
+```
+
+```
+@inproceedings{ruder-etal-2021-xtreme,
+    title = "{XTREME}-{R}: Towards More Challenging and Nuanced Multilingual Evaluation",
+    author = "Ruder, Sebastian  and
+      Constant, Noah  and
+      Botha, Jan  and
+      Siddhant, Aditya  and
+      Firat, Orhan  and
+      Fu, Jinlan  and
+      Liu, Pengfei  and
+      Hu, Junjie  and
+      Garrette, Dan  and
+      Neubig, Graham  and
+      Johnson, Melvin",
+    booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
+    month = nov,
+    year = "2021",
+    address = "Online and Punta Cana, Dominican Republic",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/2021.emnlp-main.802",
+    doi = "10.18653/v1/2021.emnlp-main.802",
+    pages = "10215--10245",
+
+}
+
+}
+```

dataset_infos.json

@@ -0,0 +1 @@
+{"ar": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "ar", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 15733091, "num_examples": 14805, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 996055, "num_examples": 921, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 3796109, "num_examples": 3661, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 1078536, "num_examples": 1594, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/arabic-train.jsonl": {"num_bytes": 18354740, "checksum": "674aeb3cdc9202c4b7359a8df16d27f57153680f7d69a6bce54fffc102c33010"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/arabic-dev.jsonl": {"num_bytes": 1230452, "checksum": "4f83dd49217fddca0cc9d3c5167146fbecf4dc075179e478ab19a2d1d4435e10"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-ar.json": {"num_bytes": 10341382, "checksum": "80afeb07f3794f8740d580ec2b27f44149e431ba36696ef86d175c2add6a53dd"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.ar-en.json": {"num_bytes": 1277232, "checksum": "5af382ecb05385bffa576bde8b5030d59fa3485105ae800067c32dd4eb804a37"}}, "download_size": 31203806, "post_processing_size": null, "dataset_size": 21603791, "size_in_bytes": 52807597}, "bn": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "bn", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4516015, "num_examples": 2390, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 223512, "num_examples": 113, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 6300125, "num_examples": 3585, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 131609, "num_examples": 181, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/bengali-train.jsonl": {"num_bytes": 4976496, "checksum": "cd9b920cc8565ec443d62b9e90252c820f8ea8f2a8516f411ae25a84468386b9"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/bengali-dev.jsonl": {"num_bytes": 254180, "checksum": "a80de94ecd8cde2189f27ea812798aa43a8456f00351cae74924ad257aa02037"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-bn.json": {"num_bytes": 12416126, "checksum": "4cfbb849d550f69145789b291511c2a78320d96499c59062659148d8d4552754"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.bn-en.json": {"num_bytes": 153978, "checksum": "8b790e3bd7abdd6196a71383a25eeee20c0d48cb96decc04c3dde81cc84f8533"}}, "download_size": 17800780, "post_processing_size": null, "dataset_size": 11171261, "size_in_bytes": 28972041}, "fi": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "fi", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 4186282, "num_examples": 6855, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 557108, "num_examples": 782, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 2733693, "num_examples": 3670, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 756233, "num_examples": 1301, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/finnish-train.jsonl": {"num_bytes": 5329446, "checksum": "e5eb49dd17070ce8776d4dd6d22bda6027e1029d02f5350b11436ad3f8e95d12"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/finnish-dev.jsonl": {"num_bytes": 743895, "checksum": "8f5f7e27c3796c060d4b7e793cda571ef4023edaae21084bd641f204c84aac95"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-fi.json": {"num_bytes": 3452924, "checksum": "a6dd3caaa2118cd8975a1db288da90dc34d96a76ad5a1b7b0d9f521a2feb5726"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.fi-en.json": {"num_bytes": 920529, "checksum": "50235118f2f179e1ad5cb7fbbda532c0e0235bbd5d05af94bc2e65c537f8bbed"}}, "download_size": 10446794, "post_processing_size": null, "dataset_size": 8233316, "size_in_bytes": 18680110}, "id": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "id", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3868663, "num_examples": 5702, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 422761, "num_examples": 565, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 2636497, "num_examples": 3667, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 581099, "num_examples": 925, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/indonesian-train.jsonl": {"num_bytes": 4837020, "checksum": "c959a3830b677b9aa18c88ba3dbbd4e4f50e06d3579951e2f0e987e629d8072b"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/indonesian-dev.jsonl": {"num_bytes": 561052, "checksum": "711c7bcceb879560867801fcec71577bee68b0d06ffcc7bc168dd1a57c13ba71"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-id.json": {"num_bytes": 3054899, "checksum": "4342aa3beabba5b014f70dd0ccf9ca496ade85eae28c260032db6ad6a7ab22e6"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.id-en.json": {"num_bytes": 702384, "checksum": "21244c217eadfa0a47a8f8f3eb68fd46745b7f762d0f187bd667762f3458d1d0"}}, "download_size": 9155355, "post_processing_size": null, "dataset_size": 7509020, "size_in_bytes": 16664375}, "ko": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "ko", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"test": {"name": "test", "num_bytes": 218745, "num_examples": 276, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 2804022, "num_examples": 3607, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 250788, "num_examples": 411, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/korean-dev.jsonl": {"num_bytes": 284268, "checksum": "8fdeb7734d3dabb90f4597e9265886995b4d3678ec7d26c76b517e9ff3d3cd4c"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-ko.json": {"num_bytes": 5274729, "checksum": "e3b92339564a02095da6eb0641ca107a2ee38c53703158fb0eb0d3bec496ddfd"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.ko-en.json": {"num_bytes": 302394, "checksum": "9c22709cd9517b1f50591177e072b7d1e9eed56cd33b5a37ceb4137457e9b2a6"}}, "download_size": 5861391, "post_processing_size": null, "dataset_size": 3273555, "size_in_bytes": 9134946}, "ru": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "ru", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 7793055, "num_examples": 6490, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 995262, "num_examples": 812, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 4179245, "num_examples": 3394, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 936798, "num_examples": 1437, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/russian-train.jsonl": {"num_bytes": 9007696, "checksum": "f6eb7e3460ebb18ae1c2d1afb0a682b9674e308178d22a0463fe7d8ddaa282f5"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/russian-dev.jsonl": {"num_bytes": 1210062, "checksum": "c5e18bed319cea27ab953b7a04d5f777c45fe43733590bb06c914d95dde9b6ff"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-ru.json": {"num_bytes": 11525665, "checksum": "d12fea5f5b54741e83288bba51dde29ad263863909f09336152aafc3ba51c34c"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.ru-en.json": {"num_bytes": 1119903, "checksum": "d9550dd86399386d8af3bd34df76716695ed6df729e0f085472acf87762cd758"}}, "download_size": 22863326, "post_processing_size": null, "dataset_size": 13904360, "size_in_bytes": 36767686}, "sw": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "sw", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 1216430, "num_examples": 2755, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 220448, "num_examples": 499, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 2620747, "num_examples": 3622, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 303729, "num_examples": 820, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/swahili-train.jsonl": {"num_bytes": 1674213, "checksum": "689f69402351bbb9b65910200095eba6b28142e4df1a63d2e7c67095ba36db74"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/swahili-dev.jsonl": {"num_bytes": 338358, "checksum": "25bc379190bc8040a9e2631322ecc3ae97940734f30f1f58bd452f0d57e808f0"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-sw.json": {"num_bytes": 3033277, "checksum": "791eca4d59c67827108997d4a5b4949055cd2e358ff8417d4d82081f960cc692"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.sw-en.json": {"num_bytes": 405346, "checksum": "0e44caac71cd4f2b446274219ce039109dfbcaab7458e225d48fb556251e23bc"}}, "download_size": 5451194, "post_processing_size": null, "dataset_size": 4361354, "size_in_bytes": 9812548}, "te": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "te", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 9008236, "num_examples": 5563, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 986265, "num_examples": 669, "dataset_name": "ty_di_qa"}, "translate_train": {"name": "translate_train", "num_bytes": 6470212, "num_examples": 3658, "dataset_name": "ty_di_qa"}, "translate_test": {"name": "translate_test", "num_bytes": 692566, "num_examples": 1135, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/telugu-train.jsonl": {"num_bytes": 10072178, "checksum": "d2af4ea2595b590a087694bf3afcb0767d18ebcc823a16bc24e3b96eb4aa10af"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/telugu-dev.jsonl": {"num_bytes": 1169646, "checksum": "773077c8232407923ae9936a4fcfef919efd307a8c8227a5c89a97b9f4d79aba"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-te.json": {"num_bytes": 12746941, "checksum": "a86908174a36da1d0d9ffa223900159dafbda299a1b13be7721a2822bd5a69e5"}, "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.te-en.json": {"num_bytes": 829082, "checksum": "c9a617a091562eab5e03bcc505a552fc6d6e4794c8c58d680eda6549698541b3"}}, "download_size": 24817847, "post_processing_size": null, "dataset_size": 17157279, "size_in_bytes": 41975126}, "en": {"description": "TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.\nThe languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language\nexpresses -- such that we expect models performing well on this set to generalize across a large number of the languages\nin the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic\ninformation-seeking task and avoid priming effects, questions are written by people who want to know the answer, but\ndon\u2019t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without\nthe use of translation (unlike MLQA and XQuAD).\n\nWe also include \"translate-train\" and \"translate-test\" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The \"translate-train\" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.\n", "citation": "@article{tydiqa,\ntitle   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},\nauthor  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}\nyear    = {2020},\njournal = {Transactions of the Association for Computational Linguistics}\n}\n", "homepage": "https://github.com/google-research-datasets/tydiqa", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "context": {"dtype": "string", "id": null, "_type": "Value"}, "question": {"dtype": "string", "id": null, "_type": "Value"}, "answers": {"feature": {"text": {"dtype": "string", "id": null, "_type": "Value"}, "answer_start": {"dtype": "int32", "id": null, "_type": "Value"}}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": [{"task": "question-answering-extractive", "question_column": "question", "context_column": "context", "answers_column": "answers"}], "builder_name": "ty_di_qa", "config_name": "en", "version": {"version_str": "1.1.0", "description": "", "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 2618156, "num_examples": 3696, "dataset_name": "ty_di_qa"}, "test": {"name": "test", "num_bytes": 343810, "num_examples": 440, "dataset_name": "ty_di_qa"}}, "download_checksums": {"https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/train/english-train.jsonl": {"num_bytes": 3247982, "checksum": "88fe80b0766db187173e36815ec06b6f156d5f5411082e0b19151c8ba7f17ddb"}, "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/dev/english-dev.jsonl": {"num_bytes": 448090, "checksum": "c98e0cb9d0fda7351d9a5fe61f1d25feaa4f035c316feeba522a6ae97dfaf18a"}}, "download_size": 3696072, "post_processing_size": null, "dataset_size": 2961966, "size_in_bytes": 6658038}}

dummy/ar/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:402e0a7bf0a29f21be9803a39d23d07c72165cdc558cf51019b798ba9c27751e
+size 7916

dummy/bn/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8214dd99e07d9c2e052c9fe427b1569e0a7f68e42bfea980cf9d7598ef1ac0d
+size 9718

dummy/en/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6751159762571ab1ed0f855898863d5b5d3a752eaba2b7bb63629eb92f85b765
+size 4074

dummy/fi/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bd0a7628cb7f834b52c54266c3480fcb299bfb83be6d89d10b1aea01d98a724
+size 6934

dummy/id/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46f4ddad708d0b102f4193099a75c4395c26d31fe7885e174a8af8782357be9e
+size 7682

dummy/ko/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e21a2d0e59e3fbb6f1897799a2d5db8daef621448c8360f5fb257670c1f3453a
+size 6530

dummy/ru/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96d03b1b879a918e0347fbbde1393b9e34a155d19ddb4da3f4b19fa89e1886b7
+size 9831

dummy/sw/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3366dd4657dbd8300eb24b31a00fea424e6d50a5e0acadae2580f3c417d82d8
+size 6233

dummy/te/1.1.0/dummy_data.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d6835616c18a79e2d3a19ce17bfda13b589c00e901962ac49f9021ade784973
+size 8443

tydiqa_xtreme.py

@@ -0,0 +1,195 @@
+import json
+import textwrap
+
+import datasets
+from datasets.tasks import QuestionAnsweringExtractive
+
+# TODO(tydiqa): BibTeX citation
+_CITATION = """\
+@article{tydiqa,
+title   = {TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},
+author  = {Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki}
+year    = {2020},
+journal = {Transactions of the Association for Computational Linguistics}
+}
+"""
+
+# TODO(tydiqa):
+_DESCRIPTION = """\
+TyDi QA is a question answering dataset covering 11 typologically diverse languages with 204K question-answer pairs.
+The languages of TyDi QA are diverse with regard to their typology -- the set of linguistic features that each language
+expresses -- such that we expect models performing well on this set to generalize across a large number of the languages
+in the world. It contains language phenomena that would not be found in English-only corpora. To provide a realistic
+information-seeking task and avoid priming effects, questions are written by people who want to know the answer, but
+don’t know the answer yet, (unlike SQuAD and its descendents) and the data is collected directly in each language without
+the use of translation (unlike MLQA and XQuAD).
+
+We also include "translate-train" and "translate-test" splits for each non-English languages from XTREME (Hu et al., 2020). These splits are the automatic translations from English to each target language used in the XTREME paper [https://arxiv.org/abs/2003.11080]. The "translate-train" split purposefully ignores the non-English TyDiQA-GoldP training data to simulate the transfer learning scenario where original-language data is not available and system builders must rely on labeled English data plus existing machine translation systems.
+"""
+
+_LANG = {
+    "ar": "arabic",
+    "bn": "bengali",
+    "en": "english",
+    "fi": "finnish",
+    "id": "indonesian",
+    "ko": "korean",
+    "ru": "russian",
+    "sw": "swahili",
+    "te": "telugu",
+}
+
+_URL_FORMAT = "https://huggingface.co/datasets/khalidalt/tydiqa-goldp/resolve/main/{split}/{lang}-{split}.jsonl"
+_TRANSLATE_TRAIN_URL_FORMAT = "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-train/tydiqa.translate.train.en-{lang}.json"
+_TRANSLATE_TEST_URL_FORMAT = "https://storage.googleapis.com/xtreme_translations/TyDiQA-GoldP/translate-test/tydiqa.translate.test.{lang}-en.json"
+
+_VERSION = datasets.Version("1.1.0", "")
+
+
+class TyDiQAConfig(datasets.BuilderConfig):
+    """BuilderConfig for TydiQa."""
+
+    def __init__(self, lang, **kwargs):
+        """
+
+        Args:
+            lang: string, language for the input text
+            **kwargs: keyword arguments forwarded to super.
+        """
+        super(TyDiQAConfig, self).__init__(version=_VERSION, **kwargs)
+        self.lang = lang
+
+class TyDiQA(datasets.GeneratorBasedBuilder):
+    """TyDi QA: Information-Seeking QA in Typologically Diverse Languages."""
+
+    BUILDER_CONFIGS = [
+        TyDiQAConfig(
+            name=lang,
+            lang=lang,
+            description=f"TyDiQA '{lang}' train and test splits, with machine-translated "
+                        "translate-train/translate-test splits "
+                        "from XTREME (Hu et al., 2020).",
+        ) for lang in _LANG if lang != "en"
+        ] + [
+            TyDiQAConfig(
+                name="en",
+                lang="en",
+                description="TyDiQA 'en' train and test splits.",
+            )
+        ]
+
+
+    def _info(self):
+        # TODO(tydiqa): Specifies the datasets.DatasetInfo object
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "id": datasets.Value("string"),
+                    "context": datasets.Value("string"),
+                    "question": datasets.Value("string"),
+                    "answers": datasets.features.Sequence(
+                        {
+                            "text": datasets.Value("string"),
+                            "answer_start": datasets.Value("int32"),
+                        }
+                    ),
+                }
+            ),
+            # No default supervised_keys (as we have to pass both question
+            # and context as input).
+            supervised_keys=None,
+            homepage="https://github.com/google-research-datasets/tydiqa",
+            citation=_CITATION,
+            task_templates=[
+                QuestionAnsweringExtractive(
+                    question_column="question", context_column="context", answers_column="answers"
+                )
+            ],
+        )
+        
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        # TODO(tydiqa): Downloads the data and defines the splits
+        # dl_manager is a datasets.download.DownloadManager that can be used to
+        # download and extract URLs
+        lang = self.config.lang
+
+        if lang == "en":
+            filepaths = dl_manager.download_and_extract({
+                "train": _URL_FORMAT.format(split="train", lang=_LANG[lang]),
+                "test": _URL_FORMAT.format(split="dev", lang=_LANG[lang])
+            })
+        elif lang == "ko":
+            filepaths = dl_manager.download_and_extract({
+                "test": _URL_FORMAT.format(split="dev", lang=_LANG[lang]),
+                "translate_train": _TRANSLATE_TRAIN_URL_FORMAT.format(lang=lang),
+                "translate_test": _TRANSLATE_TEST_URL_FORMAT.format(lang=lang),
+            })
+        else:
+            filepaths = dl_manager.download_and_extract({
+                "train": _URL_FORMAT.format(split="train", lang=_LANG[lang]),
+                "test": _URL_FORMAT.format(split="dev", lang=_LANG[lang]),
+                "translate_train": _TRANSLATE_TRAIN_URL_FORMAT.format(lang=lang),
+                "translate_test": _TRANSLATE_TEST_URL_FORMAT.format(lang=lang),
+            })
+
+        return [
+            datasets.SplitGenerator(
+                name=split,
+                # These kwargs will be passed to _generate_examples
+                gen_kwargs={"filepath": path},
+            ) for split, path in filepaths.items()
+        ]
+        
+    def _generate_examples(self, filepath):
+        """Yields examples."""
+        # TODO(tydiqa): Yields (key, example) tuples from the dataset
+        with open(filepath, encoding="utf-8") as f:
+            num_lines = sum(1 for line in f)
+        with open(filepath, encoding="utf-8") as f:
+            if num_lines == 1:
+                data = json.load(f)
+                id_ = 0
+                for article in data["data"]:
+                    for paragraph in article["paragraphs"]:
+                        context = paragraph["context"].strip()
+                        for qa in paragraph["qas"]:
+                            question = qa["question"].strip()
+
+                            answer_starts = [answer["answer_start"] for answer in qa["answers"]]
+                            answers = [answer["text"].strip() for answer in qa["answers"]]
+
+                            # Features currently used are "context", "question", and "answers".
+                            # Others are extracted here for the ease of future expansions.
+                            yield id_, {
+                                "context": context,
+                                "question": question,
+                                "id": id_,
+                                "answers": {
+                                    "answer_start": answer_starts,
+                                    "text": answers,
+                                },
+                            }
+                            id_ += 1
+            else:
+                id_ = 0
+                for line in f:
+                    data = json.loads(line)
+
+                    context = data["passage_text"].strip()
+                    question = data["question_text"].strip()
+                    answer_starts = [answer["start_byte"] for answer in data["answers"]]
+                    answers = [answer["text"].strip() for answer in data["answers"]]
+
+                    yield id_, {
+                        "context": context,
+                        "question": question,
+                        "id": id_,
+                        "answers": {
+                            "answer_start": answer_starts,
+                            "text": answers,
+                        },
+                    }
+                    id_ += 1