multitacred / dataset_infos.json
dfki-nlp's picture
Update dataset_infos.json
247237b
raw
history blame
157 kB
{"revised-en": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-en", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26061792, "num_examples": 68124, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5629179, "num_examples": 15509, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8389203, "num_examples": 22631, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 40080174, "size_in_bytes": 40237339}, "original-en": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-en", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26061792, "num_examples": 68124, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5629179, "num_examples": 15509, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8389203, "num_examples": 22631, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 40080174, "size_in_bytes": 40080174}, "original-de": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-de", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27792020, "num_examples": 67205, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6043815, "num_examples": 15282, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9007367, "num_examples": 22343, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 42843202, "size_in_bytes": 42843202}, "revised-de": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-de", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27792020, "num_examples": 67205, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6043815, "num_examples": 15282, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9007367, "num_examples": 22343, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 42843202, "size_in_bytes": 43000367}, "revised-es": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-es", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27586822, "num_examples": 65247, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5941821, "num_examples": 14908, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8921047, "num_examples": 21697, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 42449690, "size_in_bytes": 42606855}, "original-es": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-es", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 27586822, "num_examples": 65247, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5941821, "num_examples": 14908, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8921047, "num_examples": 21697, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 42449690, "size_in_bytes": 42449690}, "original-ar": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-ar", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 32371641, "num_examples": 67736, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6895001, "num_examples": 15425, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 10353930, "num_examples": 22502, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 49620572, "size_in_bytes": 49620572}, "revised-ar": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-ar", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 32371641, "num_examples": 67736, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6895001, "num_examples": 15425, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 10353930, "num_examples": 22502, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 49620572, "size_in_bytes": 49777737}, "original-fi": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-fi", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 25394979, "num_examples": 66751, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5478260, "num_examples": 15083, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8205629, "num_examples": 22268, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 39078868, "size_in_bytes": 39078868}, "revised-fi": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-fi", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 25394979, "num_examples": 66751, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5478260, "num_examples": 15083, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8205629, "num_examples": 22268, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 39078868, "size_in_bytes": 39236033}, "original-fr": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 29580179, "num_examples": 66856, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6409145, "num_examples": 15237, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9580581, "num_examples": 22248, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 45569905, "size_in_bytes": 45569905}, "revised-fr": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-fr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 29580179, "num_examples": 66856, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6409145, "num_examples": 15237, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9580581, "num_examples": 22248, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 45569905, "size_in_bytes": 45727070}, "original-hi": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-hi", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 47358490, "num_examples": 67751, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 10235547, "num_examples": 15440, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 15362616, "num_examples": 22511, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 72956653, "size_in_bytes": 72956653}, "revised-hi": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-hi", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 47358490, "num_examples": 67751, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 10235547, "num_examples": 15440, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 15362616, "num_examples": 22511, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 72956653, "size_in_bytes": 73113818}, "original-hu": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-hu", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26869925, "num_examples": 67766, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5810768, "num_examples": 15436, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8658082, "num_examples": 22519, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 41338775, "size_in_bytes": 41338775}, "revised-hu": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-hu", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26869925, "num_examples": 67766, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5810768, "num_examples": 15436, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8658082, "num_examples": 22519, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 41338775, "size_in_bytes": 41495940}, "original-ja": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-ja", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 31425001, "num_examples": 61571, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6560885, "num_examples": 13701, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9996196, "num_examples": 20290, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 47982082, "size_in_bytes": 47982082}, "revised-ja": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-ja", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 31425001, "num_examples": 61571, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 6560885, "num_examples": 13701, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 9996196, "num_examples": 20290, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 47982082, "size_in_bytes": 48139247}, "original-pl": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-pl", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26989666, "num_examples": 68124, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5845988, "num_examples": 15509, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8728082, "num_examples": 22631, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 41563736, "size_in_bytes": 41563736}, "revised-pl": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-pl", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26989666, "num_examples": 68124, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5845988, "num_examples": 15509, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8728082, "num_examples": 22631, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 41563736, "size_in_bytes": 41720901}, "original-ru": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 36546830, "num_examples": 66413, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 7846828, "num_examples": 14995, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 11847712, "num_examples": 21998, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 56241370, "size_in_bytes": 56241370}, "revised-ru": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-ru", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 36546830, "num_examples": 66413, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 7846828, "num_examples": 14995, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 11847712, "num_examples": 21998, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 56241370, "size_in_bytes": 56398535}, "original-tr": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-tr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26053517, "num_examples": 67652, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5633846, "num_examples": 15429, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8403271, "num_examples": 22510, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 40090634, "size_in_bytes": 40090634}, "revised-tr": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-tr", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26053517, "num_examples": 67652, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5633846, "num_examples": 15429, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8403271, "num_examples": 22510, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 40090634, "size_in_bytes": 40247799}, "original-zh": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "original-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26139661, "num_examples": 65211, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5483795, "num_examples": 14694, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8330520, "num_examples": 21490, "dataset_name": "multilingual_tacred"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 39953976, "size_in_bytes": 39953976}, "revised-zh": {"description": "TACRED is a large-scale relation extraction dataset with 106,264 examples built over newswire\n and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n Examples in TACRED cover 41 relation types as used in the TAC KBP challenges (e.g., per:schools_attended\n and org:members) or are labeled as no_relation if no defined relation is held. These examples are created\n by combining available human annotations from the TAC KBP challenges and crowdsourcing.\n\n Please see our EMNLP paper, or our EMNLP slides for full details.\n\nNote: There is currently a label-corrected version of the TACRED dataset, which you should consider using instead of\nthe original version released in 2017. For more details on this new version, see the TACRED Revisited paper\npublished at ACL 2020.\n\nNOTE: This Datasetreader supports a reduced version of the original TACRED JSON format with the following changes:\n- Removed fields: stanford_pos, stanford_ner, stanford_head, stanford_deprel, docid\nThe motivation for this is that we want to support additional languages, for which these fields were not required\nor available. The reader expects the specification of a language-specific configuration specifying the variant\n(original or revised) and the language (as a two-letter iso code). The default config is 'original-en'.\n\nThe Datasetreader changes the offsets of the following fields, to conform with standard Python usage (see\n#_generate_examples()):\n- subj_end to subj_end + 1 (make end offset exclusive)\n- obj_end to obj_end + 1 (make end offset exclusive)\n", "citation": "@inproceedings{zhang-etal-2017-position,\n title = \"Position-aware Attention and Supervised Data Improve Slot Filling\",\n author = \"Zhang, Yuhao and\n Zhong, Victor and\n Chen, Danqi and\n Angeli, Gabor and\n Manning, Christopher D.\",\n booktitle = \"Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing\",\n month = sep,\n year = \"2017\",\n address = \"Copenhagen, Denmark\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/D17-1004\",\n doi = \"10.18653/v1/D17-1004\",\n pages = \"35--45\",\n}\n\n@inproceedings{alt-etal-2020-tacred,\n title = \"{TACRED} Revisited: A Thorough Evaluation of the {TACRED} Relation Extraction Task\",\n author = \"Alt, Christoph and\n Gabryszak, Aleksandra and\n Hennig, Leonhard\",\n booktitle = \"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics\",\n month = jul,\n year = \"2020\",\n address = \"Online\",\n publisher = \"Association for Computational Linguistics\",\n url = \"https://www.aclweb.org/anthology/2020.acl-main.142\",\n doi = \"10.18653/v1/2020.acl-main.142\",\n pages = \"1558--1569\",\n}\n", "homepage": "https://nlp.stanford.edu/projects/tacred/", "license": "LDC", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "token": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "subj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "subj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "obj_start": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_end": {"dtype": "int32", "id": null, "_type": "Value"}, "obj_type": {"num_classes": 24, "names": ["LOCATION", "ORGANIZATION", "PERSON", "DATE", "MONEY", "PERCENT", "TIME", "CAUSE_OF_DEATH", "CITY", "COUNTRY", "CRIMINAL_CHARGE", "EMAIL", "HANDLE", "IDEOLOGY", "NATIONALITY", "RELIGION", "STATE_OR_PROVINCE", "TITLE", "URL", "NUMBER", "ORDINAL", "MISC", "DURATION", "O"], "id": null, "_type": "ClassLabel"}, "relation": {"num_classes": 42, "names": ["no_relation", "org:alternate_names", "org:city_of_headquarters", "org:country_of_headquarters", "org:dissolved", "org:founded", "org:founded_by", "org:member_of", "org:members", "org:number_of_employees/members", "org:parents", "org:political/religious_affiliation", "org:shareholders", "org:stateorprovince_of_headquarters", "org:subsidiaries", "org:top_members/employees", "org:website", "per:age", "per:alternate_names", "per:cause_of_death", "per:charges", "per:children", "per:cities_of_residence", "per:city_of_birth", "per:city_of_death", "per:countries_of_residence", "per:country_of_birth", "per:country_of_death", "per:date_of_birth", "per:date_of_death", "per:employee_of", "per:origin", "per:other_family", "per:parents", "per:religion", "per:schools_attended", "per:siblings", "per:spouse", "per:stateorprovince_of_birth", "per:stateorprovince_of_death", "per:stateorprovinces_of_residence", "per:title"], "id": null, "_type": "ClassLabel"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "multilingual_tacred", "config_name": "revised-zh", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 26139661, "num_examples": 65211, "dataset_name": "multilingual_tacred"}, "test": {"name": "test", "num_bytes": 5483795, "num_examples": 14694, "dataset_name": "multilingual_tacred"}, "validation": {"name": "validation", "num_bytes": 8330520, "num_examples": 21490, "dataset_name": "multilingual_tacred"}}, "download_checksums": {"https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/dev_patch.json": {"num_bytes": 98978, "checksum": "99002976e733e459d6d11ad45464be4e8cc5cbfe4233bca716a2fad66ee421a4"}, "https://raw.githubusercontent.com/DFKI-NLP/tacrev/master/patch/test_patch.json": {"num_bytes": 58187, "checksum": "f8816ca47c363f29e37ff5c78186a5a737704e4a14d0dd897023a6e7ab02933f"}}, "download_size": 157165, "post_processing_size": null, "dataset_size": 39953976, "size_in_bytes": 40111141}}