sucx3_ner / dataset_infos.json
Robin Kurtz
new dataset infos
2fd939f
{"original_cased": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nmanually annotated & cased", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "original_cased", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 16688117, "num_examples": 46026, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 4154574, "num_examples": 11506, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 5222628, "num_examples": 14383, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/original_tags/cased.tar.gz": {"num_bytes": 6250353, "checksum": "d365e0e4f1b2cb7221f0810152d5b33993a3221c3973cb7cf0976bcf530f5a4f"}}, "download_size": 6250353, "post_processing_size": null, "dataset_size": 26065319, "size_in_bytes": 32315672}, "original_lower": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nmanually annotated & lower", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "original_lower", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 16964273, "num_examples": 46026, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 4223610, "num_examples": 11506, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 5308926, "num_examples": 14383, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/original_tags/lower.tar.gz": {"num_bytes": 6174754, "checksum": "5930a0d6ad762c678b8eeb208155ac9b36334627d98bcc776afa6b589b4be3b5"}}, "download_size": 6174754, "post_processing_size": null, "dataset_size": 26496809, "size_in_bytes": 32671563}, "original_lower_mix": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nmanually annotated & lower_mix", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "original_lower_mix", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 16825169, "num_examples": 46026, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 4188876, "num_examples": 11506, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 5265882, "num_examples": 14383, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/original_tags/lower_mix.tar.gz": {"num_bytes": 6241014, "checksum": "bb43b2983cf7c0c5c006c78c614c3170f2bc706998c2c3336e5c364da9573b40"}}, "download_size": 6241014, "post_processing_size": null, "dataset_size": 26279927, "size_in_bytes": 32520941}, "simple_cased": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nautomatically annotated & cased", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "simple_cased", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 15227354, "num_examples": 43126, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 3793133, "num_examples": 10772, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 4787657, "num_examples": 13504, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/simple_tags/cased.tar.gz": {"num_bytes": 5147751, "checksum": "a17962cd708ab2e9dbdc7aa950487c1e791af02b2e8e6425fcfebbaf8937ebc1"}}, "download_size": 5147751, "post_processing_size": null, "dataset_size": 23808144, "size_in_bytes": 28955895}, "simple_lower": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nautomatically annotated & lower", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "simple_lower", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 15486110, "num_examples": 43126, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 3857765, "num_examples": 10772, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 4868681, "num_examples": 13504, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/simple_tags/lower.tar.gz": {"num_bytes": 5094473, "checksum": "5a52c52115e97dc288c9b17a84d364de0c9b7e3d8ee1857b3f6365eeb46c6ca1"}}, "download_size": 5094473, "post_processing_size": null, "dataset_size": 24212556, "size_in_bytes": 29307029}, "simple_lower_mix": {"description": " The dataset is a conversion of the venerable SUC 3.0 dataset into the\n huggingface ecosystem. The original dataset does not contain an official\n train-dev-test split, which is introduced here; the tag distribution for the\n NER tags between the three splits is mostly the same.\n \n The dataset has three different types of tagsets: manually annotated POS,\n manually annotated NER, and automatically annotated NER. For the\n automatically annotated NER tags, only sentences were chosen, where the\n automatic and manual annotations would match (with their respective\n categories).\n \n Additionally we provide remixes of the same data with some or all sentences\n being lowercased.\nautimatically annotated & lower_mix", "citation": "@article{gustafson2006documentation,\n title={Documentation of the Stockholm-Ume{\u0007a} Corpus},\n author={Gustafson-Capkov{'a}, Sofia and Hartmann, Britt},\n journal={Stockholm University: Department of Linguistics},\n year={2006}\n}\n", "homepage": "https://spraakbanken.gu.se/en/resources/suc3", "license": "", "features": {"id": {"dtype": "string", "id": null, "_type": "Value"}, "tokens": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "pos_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}, "ner_tags": {"feature": {"dtype": "string", "id": null, "_type": "Value"}, "length": -1, "id": null, "_type": "Sequence"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "suc3", "config_name": "simple_lower_mix", "version": {"version_str": "1.0.2", "description": null, "major": 1, "minor": 0, "patch": 2}, "splits": {"train": {"name": "train", "num_bytes": 15355562, "num_examples": 43126, "dataset_name": "suc3"}, "validation": {"name": "validation", "num_bytes": 3825257, "num_examples": 10772, "dataset_name": "suc3"}, "test": {"name": "test", "num_bytes": 4828079, "num_examples": 13504, "dataset_name": "suc3"}}, "download_checksums": {"https://huggingface.co/datasets/KBLab/sucx3_ner/resolve/main/data/simple_tags/lower_mix.tar.gz": {"num_bytes": 5152711, "checksum": "0398b4a54c40342b96f804952d7b53577157938c45ac24d368871e4e8b678abe"}}, "download_size": 5152711, "post_processing_size": null, "dataset_size": 24008898, "size_in_bytes": 29161609}}