Datasets:

---

ftopal

/

huggingface-datasets

like 0

[ "region:us" ]

[]

[]

[ "# The CoreSearch Dataset\nA large-scale dataset for cross-document event coreference search</br>\n\n- Paper: Cross-document Event Coreference Search: Task, Dataset and Modeling (link-TBD)\n \n- <ins>CoreSearchV2:</ins> A cleaner version of this dataset is now available at URL", "### Languages\n\nEnglish", "## Load Dataset\nYou can read/download the dataset files following Huggingface Hub instructions.</br>\nFor example, below code will load CoreSearch DPR folder:", "### Data Splits\n- Final version of the CD event coreference search dataset<br>\n| | Train | Valid | Test | Total |\n| ----- | ------ | ----- | ---- | ---- |\n| WEC-Eng Validated Data | | | | |\n| &nbsp;&nbsp;&nbsp;&nbsp;# Clusters | 237 | 49 | 236 | 522 | \n| &nbsp;&nbsp;&nbsp;&nbsp;# Passages (with Mentions) | 1,503 | 341 | 1,266 | 3,110 |\n| # Added Destructor Passages | 922,736 | 923,376 | 923,746 | 2,769,858 |\n| # Total Passages | 924,239 | 923,717 | 925,012 | 2,772,968 |", "## License\nWe provide the following data sets under a <a href=\"URL Commons Attribution-ShareAlike 3.0 Unported License</a>. It is based on content extracted from Wikipedia that is licensed under the Creative Commons Attribution-ShareAlike 3.0 Unported License", "## Contact\nIf you have any questions please create a Github issue at <a href=\"URL/URL" ]

[ "TAGS\n#region-us \n", "# The CoreSearch Dataset\nA large-scale dataset for cross-document event coreference search</br>\n\n- Paper: Cross-document Event Coreference Search: Task, Dataset and Modeling (link-TBD)\n \n- <ins>CoreSearchV2:</ins> A cleaner version of this dataset is now available at URL", "### Languages\n\nEnglish", "## Load Dataset\nYou can read/download the dataset files following Huggingface Hub instructions.</br>\nFor example, below code will load CoreSearch DPR folder:", "### Data Splits\n- Final version of the CD event coreference search dataset<br>\n| | Train | Valid | Test | Total |\n| ----- | ------ | ----- | ---- | ---- |\n| WEC-Eng Validated Data | | | | |\n| &nbsp;&nbsp;&nbsp;&nbsp;# Clusters | 237 | 49 | 236 | 522 | \n| &nbsp;&nbsp;&nbsp;&nbsp;# Passages (with Mentions) | 1,503 | 341 | 1,266 | 3,110 |\n| # Added Destructor Passages | 922,736 | 923,376 | 923,746 | 2,769,858 |\n| # Total Passages | 924,239 | 923,717 | 925,012 | 2,772,968 |", "## License\nWe provide the following data sets under a <a href=\"URL Commons Attribution-ShareAlike 3.0 Unported License</a>. It is based on content extracted from Wikipedia that is licensed under the Creative Commons Attribution-ShareAlike 3.0 Unported License", "## Contact\nIf you have any questions please create a Github issue at <a href=\"URL/URL" ]

[ "task_categories:image-classification", "task_ids:multi-class-image-classification", "annotations_creators:human-curator", "size_categories:1K<n<10K", "source_datasets:original", "language:en", "license:mit", "region:us" ]

[]

[ "en" ]

[ "### Dataset Summary\n\n\n![](URL\n\n\nRock Glacier Detection dataset with satelital images of rock glaciers in the Chilean Andes.", "### Supported Tasks and Leaderboards\n\n\n* 'image-classification': Based on a satelitel images (from sentinel2), the goal of this task is to predict a rock glacier in the geographic area, if there any.\n* 'image-segmentation': ...", "### Languages\n\n\nSpanish\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nA sample from the image-classification training set is provided below:\n\n\nA sample from the image-segmentation training set is provided below:", "### Data Fields\n\n\nThe data instances have the following fields:\n\n\n* 'image': A 'PIL.Image.Image' object containing the image. Note that when accessing the image column: 'dataset[0][\"image\"]' the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the '\"image\"' column, *i.e.* 'dataset[0][\"image\"]' should always be preferred over 'dataset[\"image\"][0]'.\n* 'labels': an 'int' classification label.\n\n\nClass Label Mappings:", "### Data Splits\n\n\n\nDataset Creation\n----------------", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information\n\n\nConsiderations for Using the Data\n---------------------------------", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations\n\n\nAdditional Information\n----------------------", "### Dataset Curators", "### Licensing Information", "### Contributions\n\n\nThanks to..." ]

[ "TAGS\n#task_categories-image-classification #task_ids-multi-class-image-classification #annotations_creators-human-curator #size_categories-1K<n<10K #source_datasets-original #language-English #license-mit #region-us \n", "### Dataset Summary\n\n\n![](URL\n\n\nRock Glacier Detection dataset with satelital images of rock glaciers in the Chilean Andes.", "### Supported Tasks and Leaderboards\n\n\n* 'image-classification': Based on a satelitel images (from sentinel2), the goal of this task is to predict a rock glacier in the geographic area, if there any.\n* 'image-segmentation': ...", "### Languages\n\n\nSpanish\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nA sample from the image-classification training set is provided below:\n\n\nA sample from the image-segmentation training set is provided below:", "### Data Fields\n\n\nThe data instances have the following fields:\n\n\n* 'image': A 'PIL.Image.Image' object containing the image. Note that when accessing the image column: 'dataset[0][\"image\"]' the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the '\"image\"' column, *i.e.* 'dataset[0][\"image\"]' should always be preferred over 'dataset[\"image\"][0]'.\n* 'labels': an 'int' classification label.\n\n\nClass Label Mappings:", "### Data Splits\n\n\n\nDataset Creation\n----------------", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information\n\n\nConsiderations for Using the Data\n---------------------------------", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations\n\n\nAdditional Information\n----------------------", "### Dataset Curators", "### Licensing Information", "### Contributions\n\n\nThanks to..." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Question Answering\n* Model: 21iridescent/distilbert-base-uncased-finetuned-squad\n* Dataset: adversarial_qa\n* Config: adversarialQA\n* Split: validation\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @smalllotus for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Question Answering\n* Model: 21iridescent/distilbert-base-uncased-finetuned-squad\n* Dataset: adversarial_qa\n* Config: adversarialQA\n* Split: validation\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @smalllotus for evaluating this model." ]

[ "task_ids:document-question-answering", "language:en", "arxiv:2007.00398", "region:us" ]

[ "2007.00398" ]

[ "en" ]

[ "# DocVQA: A Dataset for VQA on Document Images\n\nThe DocVQA dataset can be downloaded from the challenge page in RRC portal (\"Downloads\" tab).", "## Dataset Structure\n\nThe DocVQA comprises 50, 000 questions framed on 12,767 images. The data is split randomly in an 80−10−10 ratio to train, validation and test splits.\n- Train split: 39,463 questions, 10,194 images\n- Validation split: 5,349 questions and 1,286 images\n- Test split has 5,188 questions and 1,287 images.", "## Resources and Additional Information\n- More information can be found on the challenge page and in the DocVQA paper.\n- Document images are taken from the UCSF Industry Documents Library. It consists of a mix of printed, typewritten and handwritten content. A wide variety of document types appears in this dataset including letters, memos, notes, reports etc." ]

[ "TAGS\n#task_ids-document-question-answering #language-English #arxiv-2007.00398 #region-us \n", "# DocVQA: A Dataset for VQA on Document Images\n\nThe DocVQA dataset can be downloaded from the challenge page in RRC portal (\"Downloads\" tab).", "## Dataset Structure\n\nThe DocVQA comprises 50, 000 questions framed on 12,767 images. The data is split randomly in an 80−10−10 ratio to train, validation and test splits.\n- Train split: 39,463 questions, 10,194 images\n- Validation split: 5,349 questions and 1,286 images\n- Test split has 5,188 questions and 1,287 images.", "## Resources and Additional Information\n- More information can be found on the challenge page and in the DocVQA paper.\n- Document images are taken from the UCSF Industry Documents Library. It consists of a mix of printed, typewritten and handwritten content. A wide variety of document types appears in this dataset including letters, memos, notes, reports etc." ]

[ "task_categories:text-retrieval", "task_ids:document-retrieval", "annotations_creators:expert-generated", "multilinguality:multilingual", "language:ar", "language:bn", "language:en", "language:es", "language:fa", "language:fi", "language:fr", "language:hi", "language:id", "language:ja", "language:ko", "language:ru", "language:sw", "language:te", "language:th", "language:zh", "license:apache-2.0", "arxiv:2210.09984", "region:us" ]

[ "2210.09984" ]

[ "ar", "bn", "en", "es", "fa", "fi", "fr", "hi", "id", "ja", "ko", "ru", "sw", "te", "th", "zh" ]

[]

[ "TAGS\n#task_categories-text-retrieval #task_ids-document-retrieval #annotations_creators-expert-generated #multilinguality-multilingual #language-Arabic #language-Bengali #language-English #language-Spanish #language-Persian #language-Finnish #language-French #language-Hindi #language-Indonesian #language-Japanese #language-Korean #language-Russian #language-Swahili (macrolanguage) #language-Telugu #language-Thai #language-Chinese #license-apache-2.0 #arxiv-2210.09984 #region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/exampleem\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/exampleem\n* Config: filter\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "region:us" ]

[]

[]

[]

[ "TAGS\n#region-us \n" ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplehsd\n* Config: raw\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "task_categories:automatic-speech-recognition", "annotations_creators:crowdsourced", "language_creators:crowdsourced", "multilinguality:multilingual", "source_datasets:extended|common_voice", "license:cc0-1.0", "arxiv:1912.06670", "region:us" ]

[ "1912.06670" ]

[]

[ "# Dataset Card for Common Voice Corpus 11.0", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n - How to use\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage: URL\n- Repository: URL\n- Paper: URL\n- Leaderboard: URL\n- Point of Contact: Anton Lozhkov", "### Dataset Summary\n\nThe Common Voice dataset consists of a unique MP3 and corresponding text file. \nMany of the 24210 recorded hours in the dataset also include demographic metadata like age, sex, and accent \nthat can help improve the accuracy of speech recognition engines.\n\nThe dataset currently consists of 16413 validated hours in 100 languages, but more voices and languages are always added. \nTake a look at the Languages page to request a language or start contributing.", "### Supported Tasks and Leaderboards\n\nThe results for models trained on the Common Voice datasets are available via the \n Autoevaluate Leaderboard", "### Languages", "## How to use\n\nThe 'datasets' library allows you to load and pre-process your dataset in pure Python, at scale. The dataset can be downloaded and prepared in one call to your local drive by using the 'load_dataset' function. \n\nFor example, to download the Hindi config, simply specify the corresponding language config name (i.e., \"hi\" for Hindi):\n\n\nUsing the datasets library, you can also stream the dataset on-the-fly by adding a 'streaming=True' argument to the 'load_dataset' function call. Loading a dataset in streaming mode loads individual samples of the dataset at a time, rather than downloading the entire dataset to disk.\n\n\n*Bonus*: create a PyTorch dataloader directly with your own datasets (local/streamed).", "### Local", "### Streaming\n\n\n\nTo find out more about loading and preparing audio datasets, head over to URL", "### Example scripts\n\nTrain your own CTC or Seq2Seq Automatic Speech Recognition models on Common Voice 11 with 'transformers' - here.", "## Dataset Structure", "### Data Instances\n\nA typical data point comprises the 'path' to the audio file and its 'sentence'. \nAdditional fields include 'accent', 'age', 'client_id', 'up_votes', 'down_votes', 'gender', 'locale' and 'segment'.", "### Data Fields\n\n'client_id' ('string'): An id for which client (voice) made the recording\n\n'path' ('string'): The path to the audio file\n\n'audio' ('dict'): A dictionary containing the path to the downloaded audio file, the decoded audio array, and the sampling rate. Note that when accessing the audio column: 'dataset[0][\"audio\"]' the audio file is automatically decoded and resampled to 'dataset.features[\"audio\"].sampling_rate'. Decoding and resampling of a large number of audio files might take a significant amount of time. Thus it is important to first query the sample index before the '\"audio\"' column, *i.e.* 'dataset[0][\"audio\"]' should always be preferred over 'dataset[\"audio\"][0]'.\n\n'sentence' ('string'): The sentence the user was prompted to speak\n\n'up_votes' ('int64'): How many upvotes the audio file has received from reviewers\n\n'down_votes' ('int64'): How many downvotes the audio file has received from reviewers\n\n'age' ('string'): The age of the speaker (e.g. 'teens', 'twenties', 'fifties')\n\n'gender' ('string'): The gender of the speaker\n\n'accent' ('string'): Accent of the speaker\n\n'locale' ('string'): The locale of the speaker\n\n'segment' ('string'): Usually an empty field", "### Data Splits\n\nThe speech material has been subdivided into portions for dev, train, test, validated, invalidated, reported and other.\n\nThe validated data is data that has been validated with reviewers and received upvotes that the data is of high quality.\n\nThe invalidated data is data has been invalidated by reviewers\nand received downvotes indicating that the data is of low quality.\n\nThe reported data is data that has been reported, for different reasons.\n\nThe other data is data that has not yet been reviewed.\n\nThe dev, test, train are all data that has been reviewed, deemed of high quality and split into dev, test and train.", "## Data Preprocessing Recommended by Hugging Face\n\nThe following are data preprocessing steps advised by the Hugging Face team. They are accompanied by an example code snippet that shows how to put them to practice. \n\nMany examples in this dataset have trailing quotations marks, e.g _“the cat sat on the mat.“_. These trailing quotation marks do not change the actual meaning of the sentence, and it is near impossible to infer whether a sentence is a quotation or not a quotation from audio data alone. In these cases, it is advised to strip the quotation marks, leaving: _the cat sat on the mat_.\n\nIn addition, the majority of training sentences end in punctuation ( . or ? or ! ), whereas just a small proportion do not. In the dev set, almost all sentences end in punctuation. Thus, it is recommended to append a full-stop ( . ) to the end of the small number of training examples that do not end in punctuation.", "## Dataset Creation", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information\n\nThe dataset consists of people who have donated their voice online. You agree to not attempt to determine the identity of speakers in the Common Voice dataset.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nThe dataset consists of people who have donated their voice online. You agree to not attempt to determine the identity of speakers in the Common Voice dataset.", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information\n\nPublic Domain, CC-0" ]

[ "TAGS\n#task_categories-automatic-speech-recognition #annotations_creators-crowdsourced #language_creators-crowdsourced #multilinguality-multilingual #source_datasets-extended|common_voice #license-cc0-1.0 #arxiv-1912.06670 #region-us \n", "# Dataset Card for Common Voice Corpus 11.0", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n - How to use\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage: URL\n- Repository: URL\n- Paper: URL\n- Leaderboard: URL\n- Point of Contact: Anton Lozhkov", "### Dataset Summary\n\nThe Common Voice dataset consists of a unique MP3 and corresponding text file. \nMany of the 24210 recorded hours in the dataset also include demographic metadata like age, sex, and accent \nthat can help improve the accuracy of speech recognition engines.\n\nThe dataset currently consists of 16413 validated hours in 100 languages, but more voices and languages are always added. \nTake a look at the Languages page to request a language or start contributing.", "### Supported Tasks and Leaderboards\n\nThe results for models trained on the Common Voice datasets are available via the \n Autoevaluate Leaderboard", "### Languages", "## How to use\n\nThe 'datasets' library allows you to load and pre-process your dataset in pure Python, at scale. The dataset can be downloaded and prepared in one call to your local drive by using the 'load_dataset' function. \n\nFor example, to download the Hindi config, simply specify the corresponding language config name (i.e., \"hi\" for Hindi):\n\n\nUsing the datasets library, you can also stream the dataset on-the-fly by adding a 'streaming=True' argument to the 'load_dataset' function call. Loading a dataset in streaming mode loads individual samples of the dataset at a time, rather than downloading the entire dataset to disk.\n\n\n*Bonus*: create a PyTorch dataloader directly with your own datasets (local/streamed).", "### Local", "### Streaming\n\n\n\nTo find out more about loading and preparing audio datasets, head over to URL", "### Example scripts\n\nTrain your own CTC or Seq2Seq Automatic Speech Recognition models on Common Voice 11 with 'transformers' - here.", "## Dataset Structure", "### Data Instances\n\nA typical data point comprises the 'path' to the audio file and its 'sentence'. \nAdditional fields include 'accent', 'age', 'client_id', 'up_votes', 'down_votes', 'gender', 'locale' and 'segment'.", "### Data Fields\n\n'client_id' ('string'): An id for which client (voice) made the recording\n\n'path' ('string'): The path to the audio file\n\n'audio' ('dict'): A dictionary containing the path to the downloaded audio file, the decoded audio array, and the sampling rate. Note that when accessing the audio column: 'dataset[0][\"audio\"]' the audio file is automatically decoded and resampled to 'dataset.features[\"audio\"].sampling_rate'. Decoding and resampling of a large number of audio files might take a significant amount of time. Thus it is important to first query the sample index before the '\"audio\"' column, *i.e.* 'dataset[0][\"audio\"]' should always be preferred over 'dataset[\"audio\"][0]'.\n\n'sentence' ('string'): The sentence the user was prompted to speak\n\n'up_votes' ('int64'): How many upvotes the audio file has received from reviewers\n\n'down_votes' ('int64'): How many downvotes the audio file has received from reviewers\n\n'age' ('string'): The age of the speaker (e.g. 'teens', 'twenties', 'fifties')\n\n'gender' ('string'): The gender of the speaker\n\n'accent' ('string'): Accent of the speaker\n\n'locale' ('string'): The locale of the speaker\n\n'segment' ('string'): Usually an empty field", "### Data Splits\n\nThe speech material has been subdivided into portions for dev, train, test, validated, invalidated, reported and other.\n\nThe validated data is data that has been validated with reviewers and received upvotes that the data is of high quality.\n\nThe invalidated data is data has been invalidated by reviewers\nand received downvotes indicating that the data is of low quality.\n\nThe reported data is data that has been reported, for different reasons.\n\nThe other data is data that has not yet been reviewed.\n\nThe dev, test, train are all data that has been reviewed, deemed of high quality and split into dev, test and train.", "## Data Preprocessing Recommended by Hugging Face\n\nThe following are data preprocessing steps advised by the Hugging Face team. They are accompanied by an example code snippet that shows how to put them to practice. \n\nMany examples in this dataset have trailing quotations marks, e.g _“the cat sat on the mat.“_. These trailing quotation marks do not change the actual meaning of the sentence, and it is near impossible to infer whether a sentence is a quotation or not a quotation from audio data alone. In these cases, it is advised to strip the quotation marks, leaving: _the cat sat on the mat_.\n\nIn addition, the majority of training sentences end in punctuation ( . or ? or ! ), whereas just a small proportion do not. In the dev set, almost all sentences end in punctuation. Thus, it is recommended to append a full-stop ( . ) to the end of the small number of training examples that do not end in punctuation.", "## Dataset Creation", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information\n\nThe dataset consists of people who have donated their voice online. You agree to not attempt to determine the identity of speakers in the Common Voice dataset.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nThe dataset consists of people who have donated their voice online. You agree to not attempt to determine the identity of speakers in the Common Voice dataset.", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information\n\nPublic Domain, CC-0" ]

[ "task_categories:sentence-similarity", "language_creators:crowdsourced", "language_creators:machine-generated", "multilinguality:monolingual", "size_categories:100M<n<1B", "language:en", "license:other", "bio", "healthcare", "umls", "snomed", "definitions", "arxiv:2210.11892", "region:us" ]

[ "2210.11892" ]

[ "en" ]

[ "# The BioLORD Dataset (v1)\n\nThis dataset was constructed to enable training text embedding models producing similar representations for biomedical concept names and their definitions. Pairs of biomedical concepts names and descriptions of the concept are contrasted against each other, such that the model becomes able to find which names and descriptions are paired together within a batch.\n\n!URL\n\nThis dataset accompanies the BioLORD: Learning Ontological Representations from Definitions paper, accepted in the EMNLP 2022 Findings. When you use this dataset, please cite the original paper as follows:", "## Contents\nThe dataset contains 100M pairs (86M with descriptions, 14M with definitions). \n\n> #### Example of definitions:\n> - Site Training Documentation (Document type):\tDocument type described as records that verify completion of clinical trial site training for the site medical investigator and his/her staff.\n> - Arteries, Gastric (Arteries):\tArteries described as either of two arteries (left gastric and right gastric) that supply blood to the stomach and lesser curvature.\n> - Dental Materials, Cement, Zinc Phosphate (Biomedical or Dental Material):\tBiomedical or Dental Material described as cement dental materials, whose main components are phosphoric acid and zinc oxide, designed to produce a mechanical interlocking effect upon hardening inside the mouth. These cements consist of a basic powder (zinc oxide), an acidic liquid (phosphoric acid), and water that are mixed together in a viscous paste immediately before use, setting to a hard mass. Zinc phosphate cements have proper thermal and chemical resistance in the oral environment; they also should be resistant to dissolution in oral fluids. Zinc phosphate cements must be placed on a dental cavity liner or sealer to avoid pulp irritation. They are used in dentists' offices as cementing medium of inlays, crowns, bridges and orthodontic appliances (e.g., bands, brackets), as intermediate bases, or as temporary restorative materials.\n> - DTI (Diffusion weighted imaging):\tDiffusion weighted imaging described as a type of diffusion-weighted magnetic resonance imaging (DW-MRI) that maps the diffusion of water in three dimensions, the principal purpose of which is to image the white matter of the brain, specifically measuring the anisotropy, location, and orientation of the neural tracts, which can demonstrate microstructural changes or differences with neuropathology and treatment.\n> - arousal (psychic activity level):\tNervous System Physiological Phenomena described as cortical vigilance or readiness of tone, presumed to be in response to sensory stimulation via the reticular activating system.\n> #### Example of descriptions:\n> - Mesial fovea (Body Space or Junction):\tsomething which is a Region of surface of organ\n> - Thyroid associated opthalmopathies (Disease or Syndrome):\tsomething which has finding site orbit\n> - Internal fixation of bone of radius (Therapeutic or Preventive Procedure):\tSHOULDER AND ARM: SURGICAL REPAIRS, CLOSURES AND RECONSTRUCTIONS which has method Fixation - action\n> - gardnerella (Gram-variable bacterium):\tsomething which is a Gram-variable coccobacillus\n> - Hydropane (Organic Chemical):\tOrganic Chemical which is ingredient of homatropine / hydrocodone Oral Solution [Hydropane]\n> - Duane anomaly, myopathy, scoliosis syndrome (Multiple system malformation syndrome):\tScoliosis, unspecified which has finding site Nervous system structure\n\nAnother set of 20M descriptions based on the same knowledge graph serves as a development set (86M generations certainly do not exhaust the graph). However, this would not be a suitable test set. Instead, a test of time consisting of new concepts currently absent from UMLS would make more sense, but this will have to wait until enough new concepts have been added to UMLS.", "## License\nMy own contributions for this dataset are covered by the MIT license.\nHowever, given the data used to generate this dataset originates from UMLS, you will need to ensure you have proper licensing of UMLS before using this dataset. UMLS is free of charge in most countries, but you might have to create an account and report on your usage of the data yearly to keep a valid license." ]

[ "TAGS\n#task_categories-sentence-similarity #language_creators-crowdsourced #language_creators-machine-generated #multilinguality-monolingual #size_categories-100M<n<1B #language-English #license-other #bio #healthcare #umls #snomed #definitions #arxiv-2210.11892 #region-us \n", "# The BioLORD Dataset (v1)\n\nThis dataset was constructed to enable training text embedding models producing similar representations for biomedical concept names and their definitions. Pairs of biomedical concepts names and descriptions of the concept are contrasted against each other, such that the model becomes able to find which names and descriptions are paired together within a batch.\n\n!URL\n\nThis dataset accompanies the BioLORD: Learning Ontological Representations from Definitions paper, accepted in the EMNLP 2022 Findings. When you use this dataset, please cite the original paper as follows:", "## Contents\nThe dataset contains 100M pairs (86M with descriptions, 14M with definitions). \n\n> #### Example of definitions:\n> - Site Training Documentation (Document type):\tDocument type described as records that verify completion of clinical trial site training for the site medical investigator and his/her staff.\n> - Arteries, Gastric (Arteries):\tArteries described as either of two arteries (left gastric and right gastric) that supply blood to the stomach and lesser curvature.\n> - Dental Materials, Cement, Zinc Phosphate (Biomedical or Dental Material):\tBiomedical or Dental Material described as cement dental materials, whose main components are phosphoric acid and zinc oxide, designed to produce a mechanical interlocking effect upon hardening inside the mouth. These cements consist of a basic powder (zinc oxide), an acidic liquid (phosphoric acid), and water that are mixed together in a viscous paste immediately before use, setting to a hard mass. Zinc phosphate cements have proper thermal and chemical resistance in the oral environment; they also should be resistant to dissolution in oral fluids. Zinc phosphate cements must be placed on a dental cavity liner or sealer to avoid pulp irritation. They are used in dentists' offices as cementing medium of inlays, crowns, bridges and orthodontic appliances (e.g., bands, brackets), as intermediate bases, or as temporary restorative materials.\n> - DTI (Diffusion weighted imaging):\tDiffusion weighted imaging described as a type of diffusion-weighted magnetic resonance imaging (DW-MRI) that maps the diffusion of water in three dimensions, the principal purpose of which is to image the white matter of the brain, specifically measuring the anisotropy, location, and orientation of the neural tracts, which can demonstrate microstructural changes or differences with neuropathology and treatment.\n> - arousal (psychic activity level):\tNervous System Physiological Phenomena described as cortical vigilance or readiness of tone, presumed to be in response to sensory stimulation via the reticular activating system.\n> #### Example of descriptions:\n> - Mesial fovea (Body Space or Junction):\tsomething which is a Region of surface of organ\n> - Thyroid associated opthalmopathies (Disease or Syndrome):\tsomething which has finding site orbit\n> - Internal fixation of bone of radius (Therapeutic or Preventive Procedure):\tSHOULDER AND ARM: SURGICAL REPAIRS, CLOSURES AND RECONSTRUCTIONS which has method Fixation - action\n> - gardnerella (Gram-variable bacterium):\tsomething which is a Gram-variable coccobacillus\n> - Hydropane (Organic Chemical):\tOrganic Chemical which is ingredient of homatropine / hydrocodone Oral Solution [Hydropane]\n> - Duane anomaly, myopathy, scoliosis syndrome (Multiple system malformation syndrome):\tScoliosis, unspecified which has finding site Nervous system structure\n\nAnother set of 20M descriptions based on the same knowledge graph serves as a development set (86M generations certainly do not exhaust the graph). However, this would not be a suitable test set. Instead, a test of time consisting of new concepts currently absent from UMLS would make more sense, but this will have to wait until enough new concepts have been added to UMLS.", "## License\nMy own contributions for this dataset are covered by the MIT license.\nHowever, given the data used to generate this dataset originates from UMLS, you will need to ensure you have proper licensing of UMLS before using this dataset. UMLS is free of charge in most countries, but you might have to create an account and report on your usage of the data yearly to keep a valid license." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: gpt2\n* Dataset: mathemakitten/winobias_antistereotype_test\n* Config: mathemakitten--winobias_antistereotype_test\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @tomekkorbak for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: gpt2\n* Dataset: mathemakitten/winobias_antistereotype_test\n* Config: mathemakitten--winobias_antistereotype_test\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @tomekkorbak for evaluating this model." ]

[ "task_categories:summarization", "annotations_creators:found", "language_creators:found", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:unknown", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #annotations_creators-found #language_creators-found #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-unknown #region-us \n" ]

[ "task_categories:summarization", "annotations_creators:found", "language_creators:found", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:unknown", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #annotations_creators-found #language_creators-found #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-unknown #region-us \n" ]

[ "task_categories:summarization", "annotations_creators:found", "language_creators:found", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:unknown", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #annotations_creators-found #language_creators-found #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-unknown #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|other-MS^2", "source_datasets:extended|other-Cochrane", "language:en", "license:apache-2.0", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|other-MS^2 #source_datasets-extended|other-Cochrane #language-English #license-apache-2.0 #region-us \n" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:1K<n<10K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:1K<n<10K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:text-classification", "task_categories:question-answering", "task_categories:multiple-choice", "size_categories:1K<n<10K", "language:ru", "license:apache-2.0", "benchmark", "ethics", "question-answering", "reasoning", "arxiv:2210.12813", "region:us" ]

[ "2210.12813" ]

[ "ru" ]

[ "### Winograd\n\n\nThe Winograd schema challenge composes tasks with syntactic ambiguity, which can be resolved with logic and reasoning.", "##### Motivation\n\n\nThe dataset presents an extended version of a traditional Winograd challenge (Levesque et al., 2012): each sentence contains unresolved homonymy, which can be resolved based on commonsense and reasoning.\nThe Winograd scheme is extendable with the real-life sentences filtered out of the National Corpora with a set of 11 syntactic queries, extracting sentences like *\"Katya asked Masha if she...\"* (two possible references to a pronoun), *\"A change of scenery that...\"* (Noun phrase & subordinate clause with \"that\" in the same gender and number), etc.\nThe extraction pipeline can be adjusted to various languages depending on the set of ambiguous syntactic constructions possible.", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the dataset is a sentence with unresolved homonymy.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the sentence text\n* 'answer': a string with a candidate for the coreference resolution\n* 'options': a list of all the possible candidates present in the text\n* 'reference': a string containing an anaphor (a word or phrase that refers back to an earlier word or phrase)\n* 'homonymia\\_type': a float corresponding to the type of the structure with syntactic homonymy\n* 'label': an integer, either 0 or 1, indicating whether the homonymy is resolved correctly or not\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\n\n\nThe train and test sets are disjoint with respect to the sentence-candidate answer pairs but may include overlaps in individual sentences and homonymy type.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to six test variations, including the original test data and five adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* AddSent: generates extra words or a sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 804, Label Distribution: 66.3 / 33.7\nSplit: URL, Size (Original/Perturbed): 3458, Label Distribution: 58.1 / 41.9\nSplit: Train.episodes, Size (Original/Perturbed): 60, Label Distribution: 72.8 / 27.1\nSplit: Test.episodes, Size (Original/Perturbed): 976 / 5856, Label Distribution: 58.0 / 42.0\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe texts for the dataset are taken from the Russian National Corpus, the most representative and authoritative corpus of the Russian language available. The corpus includes texts from several domains, including news, fiction, and the web.", "##### Data Collection\n\n\nThe texts for the Winograd scheme problem are obtained using a semi-automatic pipeline.\n\n\nFirst, lists of 11 typical grammatical structures with syntactic homonymy (mainly case) are compiled. For example, two noun phrases with a complex subordinate:\n\n\nSecond, requests corresponding to these constructions are submitted to the search of the Russian National Corpus, or rather its sub-corpus with removed homonymy.\n\n\nThen, in the resulting 2k+ examples, homonymy is removed automatically with manual validation afterwards. Each original sentence is split into multiple examples in the binary classification format, indicating whether the homonymy is resolved correctly or not.\n\n\nSakaguchi et al. (2019) showed that the data Winograd Schema challenge might contain potential biases. We use the AFLite algorithm to filter out any potential biases in the data to make the test set more challenging for models. However, we do not guarantee that no spurious biases exist in the data.", "### RuWorldTree\n\n\nRuWorldTree is a QA dataset with multiple-choice elementary-level science questions, which evaluate the understanding of core science facts.", "##### Motivation\n\n\nThe WorldTree dataset starts the triad of the Reasoning and Knowledge tasks. The data includes the corpus of factoid utterances of various kinds, complex factoid questions and a corresponding causal chain of facts from the corpus resulting in a correct answer.\n\n\nThe WorldTree design was originally proposed in (Jansen et al., 2018).", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the datasets is a multiple-choice science question with 4 answer options.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the sentence text\n* 'answer': a string with a candidate for the coreference resolution\n* 'options': a list of all the possible candidates present in the text\n* 'reference': a string containing an anaphor (a word or phrase that refers back to an earlier word or phrase)\n* 'homonymia\\_type': a float corresponding to the type of the structure with syntactic homonymy\n* 'label': an integer, either 0 or 1, indicating whether the homonymy is resolved correctly or not\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\n\n\nWe use the same splits of data as in the original English version.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: replaces one or more choice options with a generated one", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 118, Label Distribution: 28.81 / 26.27 / 22.88 / 22.03\nSplit: URL, Size (Original/Perturbed): 633, Label Distribution: 22.1 / 27.5 / 25.6 / 24.8\nSplit: Train.episodes, Size (Original/Perturbed): 47, Label Distribution: 29.79 / 23.4 / 23.4 / 23.4\nSplit: Test.episodes, Size (Original/Perturbed): 629 / 4403, Label Distribution: 22.1 / 27.5 / 25.6 / 24.8\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe questions for the dataset are taken from the original WorldTree dataset, which was sourced from the AI2 Science Questions V2 corpus, consisting of both standardized exam questions from 12 US states, and the AI2 Science Questions Mercury dataset, a set of questions licensed from a student assessment entity.", "##### Data Collection\n\n\nThe dataset mainly consists of automatic translation of the English WorldTree Corpus and human validation and correction.", "### RuOpenBook\n\n\nRuOpenBookQA is a QA dataset with multiple-choice elementary-level science questions which probe the understanding of core science facts.", "##### Motivation\n\n\nRuOpenBookQA is mainly based on the work of (Mihaylov et al., 2018): it is a QA dataset with multiple-choice elementary-level science questions, which probe the understanding of 1k+ core science facts.\n\n\nVery similar to the pipeline of the RuWorldTree, the dataset includes a corpus of factoids, factoid questions and correct answer. Only one fact is enough to find the correct answer, so this task can be considered easier.", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the datasets is a multiple-choice science question with 4 answer options.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'ID': a string containing a unique question id\n* 'question': a string containing question text with answer options\n* 'answer': a string containing the correct answer key (A, B, C or D)\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: replaces one or more choice options with a generated one", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 2339, Label Distribution: 31.38 / 23.64 / 21.76 / 23.22\nSplit: URL, Size (Original/Perturbed): 500, Label Distribution: 25.2 / 27.6 / 22.0 / 25.2\nSplit: Train.episodes, Size (Original/Perturbed): 48, Label Distribution: 27.08 / 18.75 / 20.83 / 33.33\nSplit: Test.episodes, Size (Original/Perturbed): 500 / 3500, Label Distribution: 25.2 / 27.6 / 22.0 / 25.2\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe questions are taken from the original OpenBookQA dataset, created via multi-stage crowdsourcing and partial expert filtering.", "##### Data Collection\n\n\nThe dataset mainly consists of automatic translation of the English OpenBookQA and human validation and correction.", "### Ethics1\n\n\nEthics1 (sit ethics) dataset is created to test the knowledge of the basic concepts of morality. The task is to predict human ethical judgments about diverse text situations in a multi-label classification setting. Namely, the task requires models to identify the presence of concepts in normative ethics, such as virtue, law, moral, justice, and utilitarianism.", "##### Motivation\n\n\nThere is a multitude of approaches to evaluating ethics in machine learning. The Ethics dataset for Russian is created from scratch for the first time, relying on the design compatible with (Hendrycks et al., 2021).", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as excerpts from news articles and fiction texts.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the body of a news article or a fiction text\n* 'source': a string containing the source of the text\n* 'sit\\_virtue': an integer, either 0 or 1, indicating whether the concept of virtue is present in the text\n* 'sit\\_moral': an integer, either 0 or 1, indicating whether the concept of morality is present in the text\n* 'sit\\_law':an integer, either 0 or 1, indicating whether the concept of law is present in the text\n* 'sit\\_justice': an integer, either 0 or 1, indicating whether the concept of justice is present in the text\n* 'sit\\_util': an integer, either 0 or 1, indicating whether the concept of utilitarianism is present in the text\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 254, Label Distribution: 31.9 / 39.0 / 44.9 / 5.9 / 38.2\nSplit: URL, Size (Original/Perturbed): 1436, Label Distribution: 31.0 / 34.8 / 36.8 / 15.3 / 39.0\nSplit: Train.episodes, Size (Original/Perturbed): 59, Label Distribution: 30.51 / 38.98 / 35.59 / 6.78 / 37.29\nSplit: Test.episodes, Size (Original/Perturbed): 1000 / 7000, Label Distribution: 31.0 / 34.8 / 36.8 / 15.3 / 39.0\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data is sampled from the news and fiction sub-corpora of the Taiga corpus (Shavrina and Shapovalova, 2017).", "##### Data Collection\n\n\nThe composition of the dataset is conducted in a semi-automatic mode.\n\n\nFirst, lists of keywords are formulated, the presence of which in the texts means the commission of an ethically colored choice or act (e.g., 'kill', 'give', 'create', etc.). The collection of keywords includes the automatic collection of synonyms using the semantic similarity tools of the RusVestores project (Kutuzov and Kuzmenko, 2017).\n\n\nAfter that, we extract short texts containing these keywords.\n\n\nEach text is annotated via a Russian crowdsourcing platform Toloka. The workers were asked to answer five questions, one for each target column:\n\n\nDo you think the text…\n\n\n* virtue: is about someone's good/evil intentions?\n* moral: is about something that is actively approved or disapproved by society?\n* law: relates to something connected with law, routine, ceremonial?\n* justice: relates to karma (or the triumph of justice)?\n* util: refers to gains or losses (both material and emotional)?\n\n\nExamples with low inter-annotator agreement rates were filtered out.\n\n\nHuman annotators' submissions are collected and stored anonymously. The average hourly pay rate exceeds the hourly minimum wage in Russia. Each annotator is warned about potentially sensitive topics in data (e.g., politics, societal minorities, and religion).\nThe data collection process is subjected to the necessary quality review and the automatic annotation quality assessment using the honey-pot tasks.", "### Ethics2\n\n\nEthics2 (per ethics) dataset is created to test the knowledge of the basic concepts of morality. The task is to predict human ethical judgments about diverse text situations in a multi-label classification setting. The main objective of the task is to evaluate the positive or negative implementation of five concepts in normative with ‘yes’ and ‘no’ ratings. The included concepts are as follows: virtue, law, moral, justice, and utilitarianism.", "##### Motivation\n\n\nThere are a multitude of approaches to evaluating ethics in machine learning. The Ethics dataset for Russian is created from scratch for the first time, relying on the design compatible with (Hendrycks et al., 2021).\n\n\nOur Ethics dataset would go through community validation and discussion as it is the first ethics dataset for Russian based on the established methodology. We acknowledge that the work (Hendrycks et al., 2021) has flaws; thus, we do not reproduce the generative approach. We construct the dataset using a similar annotation scheme: we avoid the direct question of whether the deed is good or bad. Instead, we make annotations according to five criteria that describe the aspects of the annotators' attitude to the deed.", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as excerpts from news articles and fiction texts.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the body of a news article or a fiction text\n* 'source': a string containing the source of the text\n* 'per\\_virtue': an integer, either 0 or 1, indicating whether virtue standards are violated in the text\n* 'per\\_moral': an integer, either 0 or 1, indicating whether moral standards are violated in the text\n* 'per\\_law': an integer, either 0 or 1, indicating whether any laws are violated in the text\n* 'per\\_justice': an integer, either 0 or 1, indicating whether justice norms are violated in the text\n* 'per\\_util': an integer, either 0 or 1, indicating whether utilitarianism norms are violated in the text\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 259, Label Distribution: 69.1 / 65.3 / 78.4 / 40.9 / 23.9\nSplit: URL, Size (Original/Perturbed): 1466, Label Distribution: 64.7 / 63.5 / 78.9 / 53.0 / 27.9\nSplit: Train.episodes, Size (Original/Perturbed): 58, Label Distribution: 67.24 / 65.52 / 77.59 / 46.55 / 24.14\nSplit: Test.episodes, Size (Original/Perturbed): 1000 / 7000, Label Distribution: 64.7 / 63.5 / 78.9 / 53.0 / 27.9\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data is sampled from the news and fiction sub-corpora of the Taiga corpus (Shavrina and Shapovalova, 2017).", "##### Data Collection\n\n\nThe composition of the dataset is conducted in a semi-automatic mode.\n\n\nFirst, lists of keywords are formulated, the presence of which in the texts means the commission of an ethically colored choice or act (e.g., 'kill', 'give', 'create', etc.). The collection of keywords includes the automatic collection of synonyms using the semantic similarity tools of the RusVestores project (Kutuzov and Kuzmenko, 2017).\n\n\nAfter that, we extract short texts containing these keywords.\n\n\nEach text is annotated via a Russian crowdsourcing platform Toloka. The workers were asked to answer five questions, one for each target column:\n\n\nDo you think the text…\n\n\n* virtue: do people in the text show their best qualities or not?\n* moral: are the actions of the people in the text approved by society, regardless of their legality?\n* law: are the actions of the people in the text legal?\n* justice: do the participants receive fair retribution/reward/punishment for their deeds?\n* util: do the people in the text become wealthier/happier without making others much unhappier?\n\n\nExamples with low inter-annotator agreement rates were filtered out.\n\n\nHuman annotators' submissions are collected and stored anonymously. The average hourly pay rate exceeds the hourly minimum wage in Russia. Each annotator is warned about potentially sensitive topics in data (e.g., politics, societal minorities, and religion).\nThe data collection process is subjected to the necessary quality review and the automatic annotation quality assessment using the honey-pot tasks.", "### CheGeKa\n\n\nCheGeKa is a Jeopardy!-like Russian QA dataset collected from the official Russian quiz database ChGK.", "##### Motivation\n\n\nThe task can be considered the most challenging in terms of reasoning, knowledge and logic, as the task implies the QA pairs with a free response form (no answer choices); however, a long chain of causal relationships between facts and associations forms the correct answer.\n\n\nThe original corpus of the CheGeKa game was introduced in Mikhalkova (2021).", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as question and answer pairs.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'question\\_id': an integer corresponding to the question id in the database\n* 'question': a string containing the question text\n* 'answer': a string containing the correct answer to the question\n* 'topic': a string containing the question category\n* 'author': a string with the full name of the author\n* 'tour\\_name': a string with the title of a tournament\n* 'tour link': a string containing the link to a tournament (None for the test set)\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates extra words or a sentence at the end of the question", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split:\n\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe train data for the task was collected from the official ChGK database. Since that the database is open and its questions are easily accessed via search machines, a pack of unpublished questions written by authors of ChGK was prepared to serve as a closed test set.", "##### Data Collection\n\n\nFor information on the data collection procedure, please, refer to Mikhalkova (2021).", "### Multiq\n\n\nMultiQ is a multi-hop QA dataset for Russian, suitable for general open-domain question answering, information retrieval, and reading comprehension tasks.", "#### Motivation\n\n\nQuestion-answering has been an essential task in natural language processing and information retrieval. However, certain areas in QA remain quite challenging for modern approaches, including the multi-hop one, which is traditionally considered an intersection of graph methods, knowledge representation, and SOTA language modeling.\n\n\nMulti-hop reasoning has been the least addressed QA direction for Russian. The task is represented by the MuSeRC dataset (Fenogenova et al., 2020) and only a few dozen questions in SberQUAD (Efimov et al., 2020) and RuBQ (Rybin et al., 2021). In response, we have developed a semi-automatic pipeline for multi-hop dataset generation based on Wikidata.", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as a question with two additional texts for answer extraction.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'question': a string containing the question text\n* 'support\\_text': a string containing the first text passage relating to the question\n* 'main\\_text': a string containing the main answer text\n* 'bridge\\_answers': a list of entities required to hop from the support text to the main text\n* 'main\\_answers': a list of answers to the question\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\nTest and train data sets are disjoint with respect to individual questions, but may include overlaps in support and main texts.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split:\n\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data for the dataset is sampled from Wikipedia and Wikidata.", "##### Data Collection\n\n\nThe data for the dataset is sampled from Wikipedia and Wikidata.\n\n\nThe pipeline for dataset creation looks as follows:\n\n\nFirst, we extract the triplets from Wikidata and search for their intersections. Two triplets (subject, verb, object) are needed to compose an answerable multi-hop question. For instance, the question \"Na kakom kontinente nakhoditsya strana, grazhdaninom kotoroy byl Yokhannes Blok?\" (In what continent lies the country of which Johannes Block was a citizen?) is formed by a sequence of five graph units: \"Blok, Yokhannes\" (Block, Johannes), \"grazhdanstvo\" (country of citizenship), \"Germaniya\" (Germany), \"chast’ sveta\" (continent), and \"Yevropa\" (Europe).\n\n\nSecond, several hundreds of the question templates are curated by a few authors manually, which are further used to fine-tune ruT5-large to generate multi-hop questions given a five-fold sequence.\n\n\nThird, the resulting questions undergo paraphrasing and several rounds of manual validation procedures to control the quality and diversity.\n\n\nFinally, each question is linked to two Wikipedia paragraphs, where all graph units appear in the natural language.\n\n\nConsiderations for Using the Data\n---------------------------------", "### Societal Impact\n\n\nThe design of our benchmark allows us to alleviate the problems of a large carbon footprint (Bender et al., 2021) and keep computational costs accessible to academic and industrial fields (Couldry and Mejias, 2020). In particular, our evaluation approach does not consider LMs' fine-tuning and relies on a limited amount of episodes, while the number of attacks and perturbations can be adjusted based on the user's needs. However, achieving high robustness and task generalization may require additional computational costs based on the few-shot learning and prompting method.", "### Possible Misuse\n\n\nThe framework's usage implies working concerning zero-shot and few-shot practices, such as controlling that the test data is excluded from the pre-training corpus. Our train sets Dtrain are publicly available, and it is not anticipated that the users will apply this data for fine-tuning. Lack of control may lead to indicative and biased model evaluation.", "### Ethical Considerations\n\n\nEthics is a multidimensional subject, which remains a complicated problem for LMs and controversial for humans in a multitude of situations. Our approach is closely related to (Hendrycks et al., 2021), who introduce the ETHICS benchmark for evaluating LMs' ability to predict ethical judgments about diverse text situations. Although our methodology spans general concepts in normative ethics, we acknowledge that it can be challenging to perform objective ethical judgments about some situations (Martineau, 2006t). For instance, judgments about law are based on formal criteria (e.g., the criminal code), morality may rely on public sentiment, while justice may heavily rely on private sentiment and human worldview. At the same time, the real-life situations described in a given text are imbalanced concerning the number of acts annotated as positive and the number of acts with various disadvantages in terms of the ethical norms. In practice, this leads to the moderate inter-annotator agreement and approximate human and model performance estimates. Furthermore, other data-dependent problems can be indicated, such as genre bias and author's bias in specific publicly available text sources.\n\n\nAdditional Information\n----------------------", "### Dataset Curators\n\n\nEkaterina Taktasheva, Tatiana Shavrina, Alena Fenogenova, Denis Shevelev, Nadezhda Katricheva, Maria Tikhonova, Albina Akhmetgareeva, Oleg Zinkevich, Anastasiia Bashmakova, Svetlana Iordanskaia, Alena Spiridonova, Valentina Kurenshchikova, Ekaterina Artemova, Vladislav Mikhailov", "### Licensing Information\n\n\nApache 2.0" ]

[ "TAGS\n#task_categories-text-classification #task_categories-question-answering #task_categories-multiple-choice #size_categories-1K<n<10K #language-Russian #license-apache-2.0 #benchmark #ethics #question-answering #reasoning #arxiv-2210.12813 #region-us \n", "### Winograd\n\n\nThe Winograd schema challenge composes tasks with syntactic ambiguity, which can be resolved with logic and reasoning.", "##### Motivation\n\n\nThe dataset presents an extended version of a traditional Winograd challenge (Levesque et al., 2012): each sentence contains unresolved homonymy, which can be resolved based on commonsense and reasoning.\nThe Winograd scheme is extendable with the real-life sentences filtered out of the National Corpora with a set of 11 syntactic queries, extracting sentences like *\"Katya asked Masha if she...\"* (two possible references to a pronoun), *\"A change of scenery that...\"* (Noun phrase & subordinate clause with \"that\" in the same gender and number), etc.\nThe extraction pipeline can be adjusted to various languages depending on the set of ambiguous syntactic constructions possible.", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the dataset is a sentence with unresolved homonymy.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the sentence text\n* 'answer': a string with a candidate for the coreference resolution\n* 'options': a list of all the possible candidates present in the text\n* 'reference': a string containing an anaphor (a word or phrase that refers back to an earlier word or phrase)\n* 'homonymia\\_type': a float corresponding to the type of the structure with syntactic homonymy\n* 'label': an integer, either 0 or 1, indicating whether the homonymy is resolved correctly or not\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\n\n\nThe train and test sets are disjoint with respect to the sentence-candidate answer pairs but may include overlaps in individual sentences and homonymy type.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to six test variations, including the original test data and five adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* AddSent: generates extra words or a sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 804, Label Distribution: 66.3 / 33.7\nSplit: URL, Size (Original/Perturbed): 3458, Label Distribution: 58.1 / 41.9\nSplit: Train.episodes, Size (Original/Perturbed): 60, Label Distribution: 72.8 / 27.1\nSplit: Test.episodes, Size (Original/Perturbed): 976 / 5856, Label Distribution: 58.0 / 42.0\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe texts for the dataset are taken from the Russian National Corpus, the most representative and authoritative corpus of the Russian language available. The corpus includes texts from several domains, including news, fiction, and the web.", "##### Data Collection\n\n\nThe texts for the Winograd scheme problem are obtained using a semi-automatic pipeline.\n\n\nFirst, lists of 11 typical grammatical structures with syntactic homonymy (mainly case) are compiled. For example, two noun phrases with a complex subordinate:\n\n\nSecond, requests corresponding to these constructions are submitted to the search of the Russian National Corpus, or rather its sub-corpus with removed homonymy.\n\n\nThen, in the resulting 2k+ examples, homonymy is removed automatically with manual validation afterwards. Each original sentence is split into multiple examples in the binary classification format, indicating whether the homonymy is resolved correctly or not.\n\n\nSakaguchi et al. (2019) showed that the data Winograd Schema challenge might contain potential biases. We use the AFLite algorithm to filter out any potential biases in the data to make the test set more challenging for models. However, we do not guarantee that no spurious biases exist in the data.", "### RuWorldTree\n\n\nRuWorldTree is a QA dataset with multiple-choice elementary-level science questions, which evaluate the understanding of core science facts.", "##### Motivation\n\n\nThe WorldTree dataset starts the triad of the Reasoning and Knowledge tasks. The data includes the corpus of factoid utterances of various kinds, complex factoid questions and a corresponding causal chain of facts from the corpus resulting in a correct answer.\n\n\nThe WorldTree design was originally proposed in (Jansen et al., 2018).", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the datasets is a multiple-choice science question with 4 answer options.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the sentence text\n* 'answer': a string with a candidate for the coreference resolution\n* 'options': a list of all the possible candidates present in the text\n* 'reference': a string containing an anaphor (a word or phrase that refers back to an earlier word or phrase)\n* 'homonymia\\_type': a float corresponding to the type of the structure with syntactic homonymy\n* 'label': an integer, either 0 or 1, indicating whether the homonymy is resolved correctly or not\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\n\n\nWe use the same splits of data as in the original English version.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: replaces one or more choice options with a generated one", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 118, Label Distribution: 28.81 / 26.27 / 22.88 / 22.03\nSplit: URL, Size (Original/Perturbed): 633, Label Distribution: 22.1 / 27.5 / 25.6 / 24.8\nSplit: Train.episodes, Size (Original/Perturbed): 47, Label Distribution: 29.79 / 23.4 / 23.4 / 23.4\nSplit: Test.episodes, Size (Original/Perturbed): 629 / 4403, Label Distribution: 22.1 / 27.5 / 25.6 / 24.8\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe questions for the dataset are taken from the original WorldTree dataset, which was sourced from the AI2 Science Questions V2 corpus, consisting of both standardized exam questions from 12 US states, and the AI2 Science Questions Mercury dataset, a set of questions licensed from a student assessment entity.", "##### Data Collection\n\n\nThe dataset mainly consists of automatic translation of the English WorldTree Corpus and human validation and correction.", "### RuOpenBook\n\n\nRuOpenBookQA is a QA dataset with multiple-choice elementary-level science questions which probe the understanding of core science facts.", "##### Motivation\n\n\nRuOpenBookQA is mainly based on the work of (Mihaylov et al., 2018): it is a QA dataset with multiple-choice elementary-level science questions, which probe the understanding of 1k+ core science facts.\n\n\nVery similar to the pipeline of the RuWorldTree, the dataset includes a corpus of factoids, factoid questions and correct answer. Only one fact is enough to find the correct answer, so this task can be considered easier.", "#### Dataset Composition", "##### Data Instances\n\n\nEach instance in the datasets is a multiple-choice science question with 4 answer options.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'ID': a string containing a unique question id\n* 'question': a string containing question text with answer options\n* 'answer': a string containing the correct answer key (A, B, C or D)\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: replaces one or more choice options with a generated one", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 2339, Label Distribution: 31.38 / 23.64 / 21.76 / 23.22\nSplit: URL, Size (Original/Perturbed): 500, Label Distribution: 25.2 / 27.6 / 22.0 / 25.2\nSplit: Train.episodes, Size (Original/Perturbed): 48, Label Distribution: 27.08 / 18.75 / 20.83 / 33.33\nSplit: Test.episodes, Size (Original/Perturbed): 500 / 3500, Label Distribution: 25.2 / 27.6 / 22.0 / 25.2\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe questions are taken from the original OpenBookQA dataset, created via multi-stage crowdsourcing and partial expert filtering.", "##### Data Collection\n\n\nThe dataset mainly consists of automatic translation of the English OpenBookQA and human validation and correction.", "### Ethics1\n\n\nEthics1 (sit ethics) dataset is created to test the knowledge of the basic concepts of morality. The task is to predict human ethical judgments about diverse text situations in a multi-label classification setting. Namely, the task requires models to identify the presence of concepts in normative ethics, such as virtue, law, moral, justice, and utilitarianism.", "##### Motivation\n\n\nThere is a multitude of approaches to evaluating ethics in machine learning. The Ethics dataset for Russian is created from scratch for the first time, relying on the design compatible with (Hendrycks et al., 2021).", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as excerpts from news articles and fiction texts.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the body of a news article or a fiction text\n* 'source': a string containing the source of the text\n* 'sit\\_virtue': an integer, either 0 or 1, indicating whether the concept of virtue is present in the text\n* 'sit\\_moral': an integer, either 0 or 1, indicating whether the concept of morality is present in the text\n* 'sit\\_law':an integer, either 0 or 1, indicating whether the concept of law is present in the text\n* 'sit\\_justice': an integer, either 0 or 1, indicating whether the concept of justice is present in the text\n* 'sit\\_util': an integer, either 0 or 1, indicating whether the concept of utilitarianism is present in the text\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 254, Label Distribution: 31.9 / 39.0 / 44.9 / 5.9 / 38.2\nSplit: URL, Size (Original/Perturbed): 1436, Label Distribution: 31.0 / 34.8 / 36.8 / 15.3 / 39.0\nSplit: Train.episodes, Size (Original/Perturbed): 59, Label Distribution: 30.51 / 38.98 / 35.59 / 6.78 / 37.29\nSplit: Test.episodes, Size (Original/Perturbed): 1000 / 7000, Label Distribution: 31.0 / 34.8 / 36.8 / 15.3 / 39.0\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data is sampled from the news and fiction sub-corpora of the Taiga corpus (Shavrina and Shapovalova, 2017).", "##### Data Collection\n\n\nThe composition of the dataset is conducted in a semi-automatic mode.\n\n\nFirst, lists of keywords are formulated, the presence of which in the texts means the commission of an ethically colored choice or act (e.g., 'kill', 'give', 'create', etc.). The collection of keywords includes the automatic collection of synonyms using the semantic similarity tools of the RusVestores project (Kutuzov and Kuzmenko, 2017).\n\n\nAfter that, we extract short texts containing these keywords.\n\n\nEach text is annotated via a Russian crowdsourcing platform Toloka. The workers were asked to answer five questions, one for each target column:\n\n\nDo you think the text…\n\n\n* virtue: is about someone's good/evil intentions?\n* moral: is about something that is actively approved or disapproved by society?\n* law: relates to something connected with law, routine, ceremonial?\n* justice: relates to karma (or the triumph of justice)?\n* util: refers to gains or losses (both material and emotional)?\n\n\nExamples with low inter-annotator agreement rates were filtered out.\n\n\nHuman annotators' submissions are collected and stored anonymously. The average hourly pay rate exceeds the hourly minimum wage in Russia. Each annotator is warned about potentially sensitive topics in data (e.g., politics, societal minorities, and religion).\nThe data collection process is subjected to the necessary quality review and the automatic annotation quality assessment using the honey-pot tasks.", "### Ethics2\n\n\nEthics2 (per ethics) dataset is created to test the knowledge of the basic concepts of morality. The task is to predict human ethical judgments about diverse text situations in a multi-label classification setting. The main objective of the task is to evaluate the positive or negative implementation of five concepts in normative with ‘yes’ and ‘no’ ratings. The included concepts are as follows: virtue, law, moral, justice, and utilitarianism.", "##### Motivation\n\n\nThere are a multitude of approaches to evaluating ethics in machine learning. The Ethics dataset for Russian is created from scratch for the first time, relying on the design compatible with (Hendrycks et al., 2021).\n\n\nOur Ethics dataset would go through community validation and discussion as it is the first ethics dataset for Russian based on the established methodology. We acknowledge that the work (Hendrycks et al., 2021) has flaws; thus, we do not reproduce the generative approach. We construct the dataset using a similar annotation scheme: we avoid the direct question of whether the deed is good or bad. Instead, we make annotations according to five criteria that describe the aspects of the annotators' attitude to the deed.", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as excerpts from news articles and fiction texts.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'text': a string containing the body of a news article or a fiction text\n* 'source': a string containing the source of the text\n* 'per\\_virtue': an integer, either 0 or 1, indicating whether virtue standards are violated in the text\n* 'per\\_moral': an integer, either 0 or 1, indicating whether moral standards are violated in the text\n* 'per\\_law': an integer, either 0 or 1, indicating whether any laws are violated in the text\n* 'per\\_justice': an integer, either 0 or 1, indicating whether justice norms are violated in the text\n* 'per\\_util': an integer, either 0 or 1, indicating whether utilitarianism norms are violated in the text\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split and the label distribution:\n\n\nSplit: URL, Size (Original/Perturbed): 259, Label Distribution: 69.1 / 65.3 / 78.4 / 40.9 / 23.9\nSplit: URL, Size (Original/Perturbed): 1466, Label Distribution: 64.7 / 63.5 / 78.9 / 53.0 / 27.9\nSplit: Train.episodes, Size (Original/Perturbed): 58, Label Distribution: 67.24 / 65.52 / 77.59 / 46.55 / 24.14\nSplit: Test.episodes, Size (Original/Perturbed): 1000 / 7000, Label Distribution: 64.7 / 63.5 / 78.9 / 53.0 / 27.9\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data is sampled from the news and fiction sub-corpora of the Taiga corpus (Shavrina and Shapovalova, 2017).", "##### Data Collection\n\n\nThe composition of the dataset is conducted in a semi-automatic mode.\n\n\nFirst, lists of keywords are formulated, the presence of which in the texts means the commission of an ethically colored choice or act (e.g., 'kill', 'give', 'create', etc.). The collection of keywords includes the automatic collection of synonyms using the semantic similarity tools of the RusVestores project (Kutuzov and Kuzmenko, 2017).\n\n\nAfter that, we extract short texts containing these keywords.\n\n\nEach text is annotated via a Russian crowdsourcing platform Toloka. The workers were asked to answer five questions, one for each target column:\n\n\nDo you think the text…\n\n\n* virtue: do people in the text show their best qualities or not?\n* moral: are the actions of the people in the text approved by society, regardless of their legality?\n* law: are the actions of the people in the text legal?\n* justice: do the participants receive fair retribution/reward/punishment for their deeds?\n* util: do the people in the text become wealthier/happier without making others much unhappier?\n\n\nExamples with low inter-annotator agreement rates were filtered out.\n\n\nHuman annotators' submissions are collected and stored anonymously. The average hourly pay rate exceeds the hourly minimum wage in Russia. Each annotator is warned about potentially sensitive topics in data (e.g., politics, societal minorities, and religion).\nThe data collection process is subjected to the necessary quality review and the automatic annotation quality assessment using the honey-pot tasks.", "### CheGeKa\n\n\nCheGeKa is a Jeopardy!-like Russian QA dataset collected from the official Russian quiz database ChGK.", "##### Motivation\n\n\nThe task can be considered the most challenging in terms of reasoning, knowledge and logic, as the task implies the QA pairs with a free response form (no answer choices); however, a long chain of causal relationships between facts and associations forms the correct answer.\n\n\nThe original corpus of the CheGeKa game was introduced in Mikhalkova (2021).", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as question and answer pairs.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'question\\_id': an integer corresponding to the question id in the database\n* 'question': a string containing the question text\n* 'answer': a string containing the correct answer to the question\n* 'topic': a string containing the question category\n* 'author': a string with the full name of the author\n* 'tour\\_name': a string with the title of a tournament\n* 'tour link': a string containing the link to a tournament (None for the test set)\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates extra words or a sentence at the end of the question", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split:\n\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe train data for the task was collected from the official ChGK database. Since that the database is open and its questions are easily accessed via search machines, a pack of unpublished questions written by authors of ChGK was prepared to serve as a closed test set.", "##### Data Collection\n\n\nFor information on the data collection procedure, please, refer to Mikhalkova (2021).", "### Multiq\n\n\nMultiQ is a multi-hop QA dataset for Russian, suitable for general open-domain question answering, information retrieval, and reading comprehension tasks.", "#### Motivation\n\n\nQuestion-answering has been an essential task in natural language processing and information retrieval. However, certain areas in QA remain quite challenging for modern approaches, including the multi-hop one, which is traditionally considered an intersection of graph methods, knowledge representation, and SOTA language modeling.\n\n\nMulti-hop reasoning has been the least addressed QA direction for Russian. The task is represented by the MuSeRC dataset (Fenogenova et al., 2020) and only a few dozen questions in SberQUAD (Efimov et al., 2020) and RuBQ (Rybin et al., 2021). In response, we have developed a semi-automatic pipeline for multi-hop dataset generation based on Wikidata.", "#### Dataset Composition", "##### Data Instances\n\n\nData instances are given as a question with two additional texts for answer extraction.\n\n\nAn example in English for illustration purposes:", "##### Data Fields\n\n\n* 'question': a string containing the question text\n* 'support\\_text': a string containing the first text passage relating to the question\n* 'main\\_text': a string containing the main answer text\n* 'bridge\\_answers': a list of entities required to hop from the support text to the main text\n* 'main\\_answers': a list of answers to the question\n* 'perturbation': a string containing the name of the perturbation applied to text. If no perturbation was applied, the dataset name is used\n* 'episode': a list of episodes in which the instance is used. Only used for the train set", "##### Data Splits\n\n\nThe dataset consists of a training set with labeled examples and a test set in two configurations:\n\n\n* 'raw data': includes the original data with no additional sampling\n* 'episodes': data is split into evaluation episodes and includes several perturbations of test for robustness evaluation\nTest and train data sets are disjoint with respect to individual questions, but may include overlaps in support and main texts.", "##### Test Perturbations\n\n\nEach training episode in the dataset corresponds to seven test variations, including the original test data and six adversarial test sets, acquired through the modification of the original test through the following text perturbations:\n\n\n* ButterFingers: randomly adds noise to data by mimicking spelling mistakes made by humans through character swaps based on their keyboard distance\n* Emojify: replaces the input words with the corresponding emojis, preserving their original meaning\n* EDAdelete: randomly deletes tokens in the text\n* EDAswap: randomly swaps tokens in the text\n* BackTranslation: generates variations of the context through back-translation (ru -> en -> ru)\n* AddSent: generates an extra sentence at the end of the text", "##### General Statistics\n\n\nThe following table contains the number of examples in each data split:\n\n\n\n* 'Original' - original test data without adversarial perturbations\n* 'Perturbed' - perturbed test, containing both original data and its perturbations", "#### Dataset Creation", "##### Data Source\n\n\nThe data for the dataset is sampled from Wikipedia and Wikidata.", "##### Data Collection\n\n\nThe data for the dataset is sampled from Wikipedia and Wikidata.\n\n\nThe pipeline for dataset creation looks as follows:\n\n\nFirst, we extract the triplets from Wikidata and search for their intersections. Two triplets (subject, verb, object) are needed to compose an answerable multi-hop question. For instance, the question \"Na kakom kontinente nakhoditsya strana, grazhdaninom kotoroy byl Yokhannes Blok?\" (In what continent lies the country of which Johannes Block was a citizen?) is formed by a sequence of five graph units: \"Blok, Yokhannes\" (Block, Johannes), \"grazhdanstvo\" (country of citizenship), \"Germaniya\" (Germany), \"chast’ sveta\" (continent), and \"Yevropa\" (Europe).\n\n\nSecond, several hundreds of the question templates are curated by a few authors manually, which are further used to fine-tune ruT5-large to generate multi-hop questions given a five-fold sequence.\n\n\nThird, the resulting questions undergo paraphrasing and several rounds of manual validation procedures to control the quality and diversity.\n\n\nFinally, each question is linked to two Wikipedia paragraphs, where all graph units appear in the natural language.\n\n\nConsiderations for Using the Data\n---------------------------------", "### Societal Impact\n\n\nThe design of our benchmark allows us to alleviate the problems of a large carbon footprint (Bender et al., 2021) and keep computational costs accessible to academic and industrial fields (Couldry and Mejias, 2020). In particular, our evaluation approach does not consider LMs' fine-tuning and relies on a limited amount of episodes, while the number of attacks and perturbations can be adjusted based on the user's needs. However, achieving high robustness and task generalization may require additional computational costs based on the few-shot learning and prompting method.", "### Possible Misuse\n\n\nThe framework's usage implies working concerning zero-shot and few-shot practices, such as controlling that the test data is excluded from the pre-training corpus. Our train sets Dtrain are publicly available, and it is not anticipated that the users will apply this data for fine-tuning. Lack of control may lead to indicative and biased model evaluation.", "### Ethical Considerations\n\n\nEthics is a multidimensional subject, which remains a complicated problem for LMs and controversial for humans in a multitude of situations. Our approach is closely related to (Hendrycks et al., 2021), who introduce the ETHICS benchmark for evaluating LMs' ability to predict ethical judgments about diverse text situations. Although our methodology spans general concepts in normative ethics, we acknowledge that it can be challenging to perform objective ethical judgments about some situations (Martineau, 2006t). For instance, judgments about law are based on formal criteria (e.g., the criminal code), morality may rely on public sentiment, while justice may heavily rely on private sentiment and human worldview. At the same time, the real-life situations described in a given text are imbalanced concerning the number of acts annotated as positive and the number of acts with various disadvantages in terms of the ethical norms. In practice, this leads to the moderate inter-annotator agreement and approximate human and model performance estimates. Furthermore, other data-dependent problems can be indicated, such as genre bias and author's bias in specific publicly available text sources.\n\n\nAdditional Information\n----------------------", "### Dataset Curators\n\n\nEkaterina Taktasheva, Tatiana Shavrina, Alena Fenogenova, Denis Shevelev, Nadezhda Katricheva, Maria Tikhonova, Albina Akhmetgareeva, Oleg Zinkevich, Anastasiia Bashmakova, Svetlana Iordanskaia, Alena Spiridonova, Valentina Kurenshchikova, Ekaterina Artemova, Vladislav Mikhailov", "### Licensing Information\n\n\nApache 2.0" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:1K<n<10K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:text-to-image", "annotations_creators:machine-generated", "language_creators:other", "multilinguality:monolingual", "size_categories:n<1K", "source_datasets:N/A", "language:en", "license:cc-by-nc-sa-4.0", "region:us" ]

[]

[ "en" ]

[ "# Dataset Card for Flickr_bw_rgb\n_Dataset A image-caption dataset which stores group of black and white and color images with corresponding\n captions mentioning the content of the image with a 'colorized photograph of' or 'Black and white photograph of' suffix.\n This dataset can then be used for fine-tuning image to text models.. Only a train split is provided.", "## Examples\n \"train/<filename>.jpg\" : containing the images in JPEG format\n \"train/URL\" : Contains the metadata and the fields.\n Dataset columns:\n \"file_name\"\n \"caption\"\n\nIf you use this dataset, please cite it as:" ]

[ "TAGS\n#task_categories-text-to-image #annotations_creators-machine-generated #language_creators-other #multilinguality-monolingual #size_categories-n<1K #source_datasets-N/A #language-English #license-cc-by-nc-sa-4.0 #region-us \n", "# Dataset Card for Flickr_bw_rgb\n_Dataset A image-caption dataset which stores group of black and white and color images with corresponding\n captions mentioning the content of the image with a 'colorized photograph of' or 'Black and white photograph of' suffix.\n This dataset can then be used for fine-tuning image to text models.. Only a train split is provided.", "## Examples\n \"train/<filename>.jpg\" : containing the images in JPEG format\n \"train/URL\" : Contains the metadata and the fields.\n Dataset columns:\n \"file_name\"\n \"caption\"\n\nIf you use this dataset, please cite it as:" ]

[ "task_categories:text-classification", "task_ids:natural-language-inference", "annotations_creators:expert-generated", "language_creators:crowdsourced", "multilinguality:monolingual", "size_categories:1K<n<10K", "source_datasets:original", "language:en", "region:us" ]

[]

[ "en" ]

[ "# TRIP - Tiered Reasoning for Intuitive Physics\nOfficial dataset for Tiered Reasoning for Intuitive Physics: Toward Verifiable Commonsense Language Understanding. Shane Storks, Qiaozi Gao, Yichi Zhang, Joyce Chai. EMNLP Findings, 2021.\n\nFor our official model and experiment code, please check GitHub.", "## Overview\n!image\nWe introduce Tiered Reasoning for Intuitive Physics (TRIP), a novel commonsense reasoning dataset with dense annotations that enable multi-tiered evaluation of machines’ reasoning process.\n\nIt includes dense annotations for each story capturing multiple tiers of reasoning beyond the end task. From these annotations, we propose a tiered evaluation, where given a pair of highly similar stories (differing only by one sentence which makes one of the stories implausible), systems must jointly identify (1) the plausible story, (2) a pair of conflicting sentences in the implausible story, and (3) the underlying physical states in those sentences causing the conflict. The goal of TRIP is to enable a systematic evaluation of machine coherence toward the end task prediction of plausibility. In particular, we evaluate whether a high-level plausibility prediction can be verified based on lower-level understanding, for example, physical state changes that would support the prediction.", "## Download\n\n* HuggingFace-Dataset\n* GitHub", "## Cite" ]

[ "TAGS\n#task_categories-text-classification #task_ids-natural-language-inference #annotations_creators-expert-generated #language_creators-crowdsourced #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #region-us \n", "# TRIP - Tiered Reasoning for Intuitive Physics\nOfficial dataset for Tiered Reasoning for Intuitive Physics: Toward Verifiable Commonsense Language Understanding. Shane Storks, Qiaozi Gao, Yichi Zhang, Joyce Chai. EMNLP Findings, 2021.\n\nFor our official model and experiment code, please check GitHub.", "## Overview\n!image\nWe introduce Tiered Reasoning for Intuitive Physics (TRIP), a novel commonsense reasoning dataset with dense annotations that enable multi-tiered evaluation of machines’ reasoning process.\n\nIt includes dense annotations for each story capturing multiple tiers of reasoning beyond the end task. From these annotations, we propose a tiered evaluation, where given a pair of highly similar stories (differing only by one sentence which makes one of the stories implausible), systems must jointly identify (1) the plausible story, (2) a pair of conflicting sentences in the implausible story, and (3) the underlying physical states in those sentences causing the conflict. The goal of TRIP is to enable a systematic evaluation of machine coherence toward the end task prediction of plausibility. In particular, we evaluate whether a high-level plausibility prediction can be verified based on lower-level understanding, for example, physical state changes that would support the prediction.", "## Download\n\n* HuggingFace-Dataset\n* GitHub", "## Cite" ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Question Answering\n* Model: Adrian/distilbert-base-uncased-finetuned-squad-colab\n* Dataset: adversarial_qa\n* Config: adversarialQA\n* Split: validation\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @saad for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Question Answering\n* Model: Adrian/distilbert-base-uncased-finetuned-squad-colab\n* Dataset: adversarial_qa\n* Config: adversarialQA\n* Split: validation\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @saad for evaluating this model." ]

[ "region:us" ]

[]

[]

[]

[ "TAGS\n#region-us \n" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:other", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-other #region-us \n" ]

[ "task_categories:image-classification", "task_categories:image-to-text", "annotations_creators:crowdsourced", "language_creators:crowdsourced", "multilinguality:monolingual", "size_categories:1K<n<10K", "source_datasets:original", "language:eng", "region:us" ]

[]

[ "eng" ]

[ "# Physical-Action-Effect-Prediction\n\nOfficial dataset for \"What Action Causes This? Towards Naive Physical Action-Effect Prediction\", ACL 2018.\n\n!What Action Causes This? Towards Naive Physical Action-Effect Prediction", "## Overview\n\nDespite recent advances in knowledge representation, automated reasoning, and machine learning, artificial agents still lack the ability to understand basic action-effect relations regarding the physical world, for example, the action of cutting a cucumber most likely leads to the state where the cucumber is broken apart into smaller pieces. If artificial agents (e.g., robots) ever become our partners in joint tasks, it is critical to empower them with such action-effect understanding so that they can reason about the state of the world and plan for actions. Towards this goal, this paper introduces a new task on naive physical action-effect prediction, which addresses the relations between concrete actions (expressed in the form of verb-noun pairs) and their effects on the state of the physical world as depicted by images. We collected a dataset for this task and developed an approach that harnesses web image data through distant supervision to facilitate learning for action-effect prediction. Our empirical results have shown that web data can be used to complement a small number of seed examples (e.g., three examples for each action) for model learning. This opens up possibilities for agents to learn physical action-effect relations for tasks at hand through communication with humans with a few examples.", "### Datasets\n\n- This dataset contains action-effect information for 140 verb-noun pairs. It has two parts: effects described by natural language, and effects depicted in images.\n- The language data contains verb-noun pairs and their effects described in natural language. For each verb-noun pair, its possible effects are described by 10 different annotators. The format for each line is 'verb noun, effect_sentence, [effect_phrase_1, effect_phrase_2, effect_phrase_3, ...]'. Effect_phrases were automatically extracted from their corresponding effect_sentences. \n- The image data contains images depicting action effects. For each verb-noun pair, an average of 15 positive images and 15 negative images were collected. Positive images are those deemed to capture the resulting world state of the action. And negative images are those deemed to capture some state of the related object (*i.e.*, the nouns in the verb-noun pairs), but are not the resulting state of the corresponding action.", "### Download\n\n* HuggingFace\n* Google Drive\n* Dropbox:\n * Language Data\n * Image Data", "### Cite\n\nWhat Action Causes This? Towards Naïve Physical Action-Effect Prediction. *Qiaozi Gao, Shaohua Yang, Joyce Chai, Lucy Vanderwende*. ACL, 2018. [[Paper]](URL [[Slides]](URL" ]

[ "TAGS\n#task_categories-image-classification #task_categories-image-to-text #annotations_creators-crowdsourced #language_creators-crowdsourced #multilinguality-monolingual #size_categories-1K<n<10K #source_datasets-original #language-English #region-us \n", "# Physical-Action-Effect-Prediction\n\nOfficial dataset for \"What Action Causes This? Towards Naive Physical Action-Effect Prediction\", ACL 2018.\n\n!What Action Causes This? Towards Naive Physical Action-Effect Prediction", "## Overview\n\nDespite recent advances in knowledge representation, automated reasoning, and machine learning, artificial agents still lack the ability to understand basic action-effect relations regarding the physical world, for example, the action of cutting a cucumber most likely leads to the state where the cucumber is broken apart into smaller pieces. If artificial agents (e.g., robots) ever become our partners in joint tasks, it is critical to empower them with such action-effect understanding so that they can reason about the state of the world and plan for actions. Towards this goal, this paper introduces a new task on naive physical action-effect prediction, which addresses the relations between concrete actions (expressed in the form of verb-noun pairs) and their effects on the state of the physical world as depicted by images. We collected a dataset for this task and developed an approach that harnesses web image data through distant supervision to facilitate learning for action-effect prediction. Our empirical results have shown that web data can be used to complement a small number of seed examples (e.g., three examples for each action) for model learning. This opens up possibilities for agents to learn physical action-effect relations for tasks at hand through communication with humans with a few examples.", "### Datasets\n\n- This dataset contains action-effect information for 140 verb-noun pairs. It has two parts: effects described by natural language, and effects depicted in images.\n- The language data contains verb-noun pairs and their effects described in natural language. For each verb-noun pair, its possible effects are described by 10 different annotators. The format for each line is 'verb noun, effect_sentence, [effect_phrase_1, effect_phrase_2, effect_phrase_3, ...]'. Effect_phrases were automatically extracted from their corresponding effect_sentences. \n- The image data contains images depicting action effects. For each verb-noun pair, an average of 15 positive images and 15 negative images were collected. Positive images are those deemed to capture the resulting world state of the action. And negative images are those deemed to capture some state of the related object (*i.e.*, the nouns in the verb-noun pairs), but are not the resulting state of the corresponding action.", "### Download\n\n* HuggingFace\n* Google Drive\n* Dropbox:\n * Language Data\n * Image Data", "### Cite\n\nWhat Action Causes This? Towards Naïve Physical Action-Effect Prediction. *Qiaozi Gao, Shaohua Yang, Joyce Chai, Lucy Vanderwende*. ACL, 2018. [[Paper]](URL [[Slides]](URL" ]

[ "region:us" ]

[]

[]

[]

[ "TAGS\n#region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"corporate-surrealist-training\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"corporate-surrealist-training\"\n\nMore Information needed" ]

[ "task_categories:translation", "multilinguality:multilingual", "source_datasets:FLORES-101, FLORES-200, PAWS-X, XNLI, XTREME, WinoMT, Wino-X, MuCOW, EuroParl ConDisco, ParcorFull", "language:multilingual", "license:cc-by-nc-sa-4.0", "arxiv:2401.16313", "region:us" ]

[ "2401.16313" ]

[ "multilingual" ]

[ "# Dataset Card for ACES and Span-ACES", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Discussion of Biases\n - Usage\n - Other Known Limitations\n- Additional Information\n - Licensing Information\n - Citation Information\n - Contact", "## Dataset Description\n\n- Repository: ACES dataset repository\n- Paper: arXiv", "### Dataset Summary\n\nACES consists of 36,476 examples covering 146 language pairs and representing challenges from 68 phenomena for evaluating machine translation metrics. We focus on translation accuracy errors and base the phenomena covered in our challenge set on the Multidimensional Quality Metrics (MQM) ontology. The phenomena range from simple perturbations at the word/character level to more complex errors based on discourse and real-world knowledge. \n29.01.2024: We also release Span-ACES, which is an extension to the ACES dataset. The errors in incorrect-translation are explicitly marked in a <v>span</v> format.", "### Supported Tasks and Leaderboards\n\n-Machine translation evaluation of metrics\n\n-Potentially useful for contrastive machine translation evaluation", "### Languages\n\nThe dataset covers 146 language pairs as follows: \n\naf-en, af-fa, ar-en, ar-fr, ar-hi, be-en, bg-en, bg-lt, ca-en, ca-es, cs-en, da-en, de-en, de-es, de-fr, de-ja, de-ko, de-ru, de-zh, el-en, en-af, en-ar, en-be, en-bg, en-ca, en-cs, en-da, en-de, en-el, en-es, en-et, en-fa, en-fi, en-fr, en-gl, en-he, en-hi, en-hr, en-hu, en-hy, en-id, en-it, en-ja, en-ko, en-lt, en-lv, en-mr, en-nl, en-no, en-pl, en-pt, en-ro, en-ru, en-sk, en-sl, en-sr, en-sv, en-ta, en-tr, en-uk, en-ur, en-vi, en-zh, es-ca, es-de, es-en, es-fr, es-ja, es-ko, es-zh, et-en, fa-af, fa-en, fi-en, fr-de, fr-en, fr-es, fr-ja, fr-ko, fr-mr, fr-ru, fr-zh, ga-en, gl-en, he-en, he-sv, hi-ar, hi-en, hr-en, hr-lv, hu-en, hy-en, hy-vi, id-en, it-en, ja-de, ja-en, ja-es, ja-fr, ja-ko, ja-zh, ko-de, ko-en, ko-es, ko-fr, ko-ja, ko-zh, lt-bg, lt-en, lv-en, lv-hr, mr-en, nl-en, no-en, pl-en, pl-mr, pl-sk, pt-en, pt-sr, ro-en, ru-de, ru-en, ru-es, ru-fr, sk-en, sk-pl, sl-en, sr-en, sr-pt, sv-en, sv-he, sw-en, ta-en, th-en, tr-en, uk-en, ur-en, vi-en, vi-hy, wo-en, zh-de, zh-en, zh-es, zh-fr, zh-ja, zh-ko", "## Dataset Structure", "### Data Instances\n\nEach data instance contains the following features: _source_, _good-translation_, _incorrect-translation_, _reference_, _phenomena_, _langpair_\n\nSee the ACES corpus viewer to explore more examples.\n\nAn example from the ACES challenge set looks like the following:\n\n\nAn example from the Span-ACES challenge set looks like the following:", "### Data Fields\n\n- 'source': a string containing the text that needs to be translated\n- 'good-translation': possible translation of the source sentence\n- 'incorrect-translation': translation of the source sentence that contains an error or phenomenon of interest\n- 'reference': the gold standard translation \n- 'phenomena': the type of error or phenomena being studied in the example\n- 'langpair': the source language and the target language pair of the example\n- 'incorrect-translation-annotated': incorrect translation with annotated spans containing the phenomena\n- 'annotation-method': field describing how the annotation \n\nNote that the _good-translation_ may not be free of errors but it is a better translation than the _incorrect-translation_", "### Data Splits\n\nThe ACES dataset has 1 split: _train_ which contains the challenge set. There are 36476 examples. \nNote, the examples in Span-ACES are identical to ACES with the two additional columns. The examples are also stored under a different _train_ split", "## Dataset Creation", "### Curation Rationale\n\nWith the advent of neural networks and especially Transformer-based architectures, machine translation outputs have become more and more fluent. Fluency errors are also judged less severely than accuracy errors by human evaluators \\citep{freitag-etal-2021-experts} which reflects the fact that accuracy errors can have dangerous consequences in certain contexts, for example in the medical and legal domains. For these reasons, we decided to build a challenge set focused on accuracy errors. \n\nAnother aspect we focus on is including a broad range of language pairs in ACES. Whenever possible we create examples for all language pairs covered in a source dataset when we use automatic approaches. For phenomena where we create examples manually, we also aim to cover at least two language pairs per phenomenon but are of course limited to the languages spoken by the authors.\n\nWe aim to offer a collection of challenge sets covering both easy and hard phenomena. While it may be of interest to the community to continuously test on harder examples to check where machine translation evaluation metrics still break, we believe that easy challenge sets are just as important to ensure that metrics do not suddenly become worse at identifying error types that were previously considered ''solved''. Therefore, we take a holistic view when creating ACES and do not filter out individual examples or exclude challenge sets based on baseline metric performance or other factors.", "### Source Data", "#### Initial Data Collection and Normalization\n\nPlease see Sections 4 and 5 of the paper.", "#### Who are the source language producers?\n\nThe dataset contains sentences found in FLORES-101, FLORES-200, PAWS-X, XNLI, XTREME, WinoMT, Wino-X, MuCOW, EuroParl ConDisco, ParcorFull datasets. Please refer to the respective papers for further details.", "### Personal and Sensitive Information\n\nThe external datasets may contain sensitive information. Refer to the respective datasets for further details.", "## Considerations for Using the Data", "### Usage\n\nACES has been primarily designed to evaluate machine translation metrics on the accuracy errors. We expect the metric to score _good-translation_ consistently higher than _incorrect-translation_. We report the performance of metric based on Kendall-tau like correlation. It measures the number of times a metric scores the good translation above the incorrect translation (concordant) and equal to or lower than the incorrect translation (discordant).", "### Discussion of Biases\n\nSome examples within the challenge set exhibit biases, however, this is necessary in order to expose the limitations of existing metrics.", "### Other Known Limitations\nThe ACES challenge set exhibits a number of biases. Firstly, there is greater coverage in terms of phenomena and the number of examples for the en-de and en-fr language pairs. This is in part due to the manual effort required to construct examples for some phenomena, in particular, those belonging to the discourse-level and real-world knowledge categories. Further, our choice of language pairs is also limited to the ones available in XLM-R. Secondly, ACES contains more examples for those phenomena for which examples could be generated automatically, compared to those that required manual construction/filtering. Thirdly, some of the automatically generated examples require external libraries which are only available for a few languages (e.g. Multilingual Wordnet). Fourthly, the focus of the challenge set is on accuracy errors. We leave the development of challenge sets for fluency errors to future work.\n\nAs a result of using existing datasets as the basis for many of the examples, errors present in these datasets may be propagated through into ACES. Whilst we acknowledge that this is undesirable, in our methods for constructing the incorrect translation we aim to ensure that the quality of the incorrect translation is always worse than the corresponding good translation.\n\nThe results and analyses presented in the paper exclude those metrics submitted to the WMT 2022 metrics shared task that provides only system-level outputs. We focus on metrics that provide segment-level outputs as this enables us to provide a broad overview of metric performance on different phenomenon categories and to conduct fine-grained analyses of performance on individual phenomena. For some of the fine-grained analyses, we apply additional constraints based on the language pairs covered by the metrics, or whether the metrics take the source as input, to address specific questions of interest. As a result of applying some of these additional constraints, our investigations tend to focus more on high and medium-resource languages than on low-resource languages. We hope to address this shortcoming in future work.", "## Additional Information", "### Licensing Information\n\nThe ACES dataset is Creative Commons Attribution Non-Commercial Share Alike 4.0 (cc-by-nc-sa-4.0) \n\n\n\n\n\n\nIf using Span-ACES,", "### Contact\nChantal Amrhein and Nikita Moghe and Liane Guillou\n\nDataset card based on Allociné" ]

[ "TAGS\n#task_categories-translation #multilinguality-multilingual #source_datasets-FLORES-101, FLORES-200, PAWS-X, XNLI, XTREME, WinoMT, Wino-X, MuCOW, EuroParl ConDisco, ParcorFull #language-multilingual #license-cc-by-nc-sa-4.0 #arxiv-2401.16313 #region-us \n", "# Dataset Card for ACES and Span-ACES", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Discussion of Biases\n - Usage\n - Other Known Limitations\n- Additional Information\n - Licensing Information\n - Citation Information\n - Contact", "## Dataset Description\n\n- Repository: ACES dataset repository\n- Paper: arXiv", "### Dataset Summary\n\nACES consists of 36,476 examples covering 146 language pairs and representing challenges from 68 phenomena for evaluating machine translation metrics. We focus on translation accuracy errors and base the phenomena covered in our challenge set on the Multidimensional Quality Metrics (MQM) ontology. The phenomena range from simple perturbations at the word/character level to more complex errors based on discourse and real-world knowledge. \n29.01.2024: We also release Span-ACES, which is an extension to the ACES dataset. The errors in incorrect-translation are explicitly marked in a <v>span</v> format.", "### Supported Tasks and Leaderboards\n\n-Machine translation evaluation of metrics\n\n-Potentially useful for contrastive machine translation evaluation", "### Languages\n\nThe dataset covers 146 language pairs as follows: \n\naf-en, af-fa, ar-en, ar-fr, ar-hi, be-en, bg-en, bg-lt, ca-en, ca-es, cs-en, da-en, de-en, de-es, de-fr, de-ja, de-ko, de-ru, de-zh, el-en, en-af, en-ar, en-be, en-bg, en-ca, en-cs, en-da, en-de, en-el, en-es, en-et, en-fa, en-fi, en-fr, en-gl, en-he, en-hi, en-hr, en-hu, en-hy, en-id, en-it, en-ja, en-ko, en-lt, en-lv, en-mr, en-nl, en-no, en-pl, en-pt, en-ro, en-ru, en-sk, en-sl, en-sr, en-sv, en-ta, en-tr, en-uk, en-ur, en-vi, en-zh, es-ca, es-de, es-en, es-fr, es-ja, es-ko, es-zh, et-en, fa-af, fa-en, fi-en, fr-de, fr-en, fr-es, fr-ja, fr-ko, fr-mr, fr-ru, fr-zh, ga-en, gl-en, he-en, he-sv, hi-ar, hi-en, hr-en, hr-lv, hu-en, hy-en, hy-vi, id-en, it-en, ja-de, ja-en, ja-es, ja-fr, ja-ko, ja-zh, ko-de, ko-en, ko-es, ko-fr, ko-ja, ko-zh, lt-bg, lt-en, lv-en, lv-hr, mr-en, nl-en, no-en, pl-en, pl-mr, pl-sk, pt-en, pt-sr, ro-en, ru-de, ru-en, ru-es, ru-fr, sk-en, sk-pl, sl-en, sr-en, sr-pt, sv-en, sv-he, sw-en, ta-en, th-en, tr-en, uk-en, ur-en, vi-en, vi-hy, wo-en, zh-de, zh-en, zh-es, zh-fr, zh-ja, zh-ko", "## Dataset Structure", "### Data Instances\n\nEach data instance contains the following features: _source_, _good-translation_, _incorrect-translation_, _reference_, _phenomena_, _langpair_\n\nSee the ACES corpus viewer to explore more examples.\n\nAn example from the ACES challenge set looks like the following:\n\n\nAn example from the Span-ACES challenge set looks like the following:", "### Data Fields\n\n- 'source': a string containing the text that needs to be translated\n- 'good-translation': possible translation of the source sentence\n- 'incorrect-translation': translation of the source sentence that contains an error or phenomenon of interest\n- 'reference': the gold standard translation \n- 'phenomena': the type of error or phenomena being studied in the example\n- 'langpair': the source language and the target language pair of the example\n- 'incorrect-translation-annotated': incorrect translation with annotated spans containing the phenomena\n- 'annotation-method': field describing how the annotation \n\nNote that the _good-translation_ may not be free of errors but it is a better translation than the _incorrect-translation_", "### Data Splits\n\nThe ACES dataset has 1 split: _train_ which contains the challenge set. There are 36476 examples. \nNote, the examples in Span-ACES are identical to ACES with the two additional columns. The examples are also stored under a different _train_ split", "## Dataset Creation", "### Curation Rationale\n\nWith the advent of neural networks and especially Transformer-based architectures, machine translation outputs have become more and more fluent. Fluency errors are also judged less severely than accuracy errors by human evaluators \\citep{freitag-etal-2021-experts} which reflects the fact that accuracy errors can have dangerous consequences in certain contexts, for example in the medical and legal domains. For these reasons, we decided to build a challenge set focused on accuracy errors. \n\nAnother aspect we focus on is including a broad range of language pairs in ACES. Whenever possible we create examples for all language pairs covered in a source dataset when we use automatic approaches. For phenomena where we create examples manually, we also aim to cover at least two language pairs per phenomenon but are of course limited to the languages spoken by the authors.\n\nWe aim to offer a collection of challenge sets covering both easy and hard phenomena. While it may be of interest to the community to continuously test on harder examples to check where machine translation evaluation metrics still break, we believe that easy challenge sets are just as important to ensure that metrics do not suddenly become worse at identifying error types that were previously considered ''solved''. Therefore, we take a holistic view when creating ACES and do not filter out individual examples or exclude challenge sets based on baseline metric performance or other factors.", "### Source Data", "#### Initial Data Collection and Normalization\n\nPlease see Sections 4 and 5 of the paper.", "#### Who are the source language producers?\n\nThe dataset contains sentences found in FLORES-101, FLORES-200, PAWS-X, XNLI, XTREME, WinoMT, Wino-X, MuCOW, EuroParl ConDisco, ParcorFull datasets. Please refer to the respective papers for further details.", "### Personal and Sensitive Information\n\nThe external datasets may contain sensitive information. Refer to the respective datasets for further details.", "## Considerations for Using the Data", "### Usage\n\nACES has been primarily designed to evaluate machine translation metrics on the accuracy errors. We expect the metric to score _good-translation_ consistently higher than _incorrect-translation_. We report the performance of metric based on Kendall-tau like correlation. It measures the number of times a metric scores the good translation above the incorrect translation (concordant) and equal to or lower than the incorrect translation (discordant).", "### Discussion of Biases\n\nSome examples within the challenge set exhibit biases, however, this is necessary in order to expose the limitations of existing metrics.", "### Other Known Limitations\nThe ACES challenge set exhibits a number of biases. Firstly, there is greater coverage in terms of phenomena and the number of examples for the en-de and en-fr language pairs. This is in part due to the manual effort required to construct examples for some phenomena, in particular, those belonging to the discourse-level and real-world knowledge categories. Further, our choice of language pairs is also limited to the ones available in XLM-R. Secondly, ACES contains more examples for those phenomena for which examples could be generated automatically, compared to those that required manual construction/filtering. Thirdly, some of the automatically generated examples require external libraries which are only available for a few languages (e.g. Multilingual Wordnet). Fourthly, the focus of the challenge set is on accuracy errors. We leave the development of challenge sets for fluency errors to future work.\n\nAs a result of using existing datasets as the basis for many of the examples, errors present in these datasets may be propagated through into ACES. Whilst we acknowledge that this is undesirable, in our methods for constructing the incorrect translation we aim to ensure that the quality of the incorrect translation is always worse than the corresponding good translation.\n\nThe results and analyses presented in the paper exclude those metrics submitted to the WMT 2022 metrics shared task that provides only system-level outputs. We focus on metrics that provide segment-level outputs as this enables us to provide a broad overview of metric performance on different phenomenon categories and to conduct fine-grained analyses of performance on individual phenomena. For some of the fine-grained analyses, we apply additional constraints based on the language pairs covered by the metrics, or whether the metrics take the source as input, to address specific questions of interest. As a result of applying some of these additional constraints, our investigations tend to focus more on high and medium-resource languages than on low-resource languages. We hope to address this shortcoming in future work.", "## Additional Information", "### Licensing Information\n\nThe ACES dataset is Creative Commons Attribution Non-Commercial Share Alike 4.0 (cc-by-nc-sa-4.0) \n\n\n\n\n\n\nIf using Span-ACES,", "### Contact\nChantal Amrhein and Nikita Moghe and Liane Guillou\n\nDataset card based on Allociné" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"leaflet_offers\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"leaflet_offers\"\n\nMore Information needed" ]

[ "task_categories:summarization", "task_ids:news-articles-summarization", "annotations_creators:other", "language_creators:other", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:extended|orange_sum", "language:fr", "license:other", "krm", "ulpgl", "orange", "region:us" ]

[]

[ "fr" ]

[ "# Dataset Card for [for-ULPGL-Dissertation]", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage: krm/for-ULPGL-Dissertation\n- Repository:\n- Paper:\n- Leaderboard:\n- Point of Contact:", "### Dataset Summary\n\nCe dataset est essentiellement basé sur le dataset *GEM/Orange_sum* dédié à la synthèse d'articles en français. Il est constitué des données abstract de ce dataset (Orange_sum) auxquelles a été ajouté un certain nombre de synthèses générées par le système Mon Résumeur de David Krame.", "### Supported Tasks and Leaderboards\n\nSynthèse automatique", "### Languages\n\nFrançais", "## Dataset Structure", "### Data Fields\n\n*summary* et *text* sont les champs du dataset avec :\n\ntext contient les textes et\nsummary les synthèses correspondantes.", "### Data Splits\n\nPour le moment (le 16 Octobre 2022), le dataset est constitué de :\n\n> 21721 données d'entraînement (split dénommé train)\n\n> 1545 données de validation (split dénommé validation)\n\n> 1581 données de test (split dénommé test)", "## Dataset Creation", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information", "## Considerations for Using the Data", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information", "### Contributions" ]

[ "TAGS\n#task_categories-summarization #task_ids-news-articles-summarization #annotations_creators-other #language_creators-other #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-extended|orange_sum #language-French #license-other #krm #ulpgl #orange #region-us \n", "# Dataset Card for [for-ULPGL-Dissertation]", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage: krm/for-ULPGL-Dissertation\n- Repository:\n- Paper:\n- Leaderboard:\n- Point of Contact:", "### Dataset Summary\n\nCe dataset est essentiellement basé sur le dataset *GEM/Orange_sum* dédié à la synthèse d'articles en français. Il est constitué des données abstract de ce dataset (Orange_sum) auxquelles a été ajouté un certain nombre de synthèses générées par le système Mon Résumeur de David Krame.", "### Supported Tasks and Leaderboards\n\nSynthèse automatique", "### Languages\n\nFrançais", "## Dataset Structure", "### Data Fields\n\n*summary* et *text* sont les champs du dataset avec :\n\ntext contient les textes et\nsummary les synthèses correspondantes.", "### Data Splits\n\nPour le moment (le 16 Octobre 2022), le dataset est constitué de :\n\n> 21721 données d'entraînement (split dénommé train)\n\n> 1545 données de validation (split dénommé validation)\n\n> 1581 données de test (split dénommé test)", "## Dataset Creation", "### Curation Rationale", "### Source Data", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?", "### Personal and Sensitive Information", "## Considerations for Using the Data", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information", "### Contributions" ]

[ "doi:10.57967/hf/0041", "region:us" ]

[]

[]

[]

[ "TAGS\n#doi-10.57967/hf/0041 #region-us \n" ]

[ "task_categories:text-classification", "task_ids:natural-language-inference", "annotations_creators:no-annotation", "language_creators:found", "multilinguality:monolingual", "size_categories:100K<n<1M", "source_datasets:snli", "source_datasets:multi_nli", "source_datasets:anli", "language:en", "license:agpl-3.0", "region:us" ]

[]

[ "en" ]

[ "# Repo\n\nGithub Repo: thamognya/TBertNLI specifically in the src/data directory.", "# Sample", "# Datsets Origin\n\nAs of now the marked datasets have been used to make this dataset and the other ones are todo\n\n- [x] SNLI\n- [x] MultiNLI\n- SuperGLUE\n- FEVER\n- WIKI-FACTCHECK\n- [x] ANLI\n- more from huggingface", "# Reasons\n\nJust for finetuning of NLI models and purely made for NLI (not zero shot classification)" ]

[ "TAGS\n#task_categories-text-classification #task_ids-natural-language-inference #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-100K<n<1M #source_datasets-snli #source_datasets-multi_nli #source_datasets-anli #language-English #license-agpl-3.0 #region-us \n", "# Repo\n\nGithub Repo: thamognya/TBertNLI specifically in the src/data directory.", "# Sample", "# Datsets Origin\n\nAs of now the marked datasets have been used to make this dataset and the other ones are todo\n\n- [x] SNLI\n- [x] MultiNLI\n- SuperGLUE\n- FEVER\n- WIKI-FACTCHECK\n- [x] ANLI\n- more from huggingface", "# Reasons\n\nJust for finetuning of NLI models and purely made for NLI (not zero shot classification)" ]

[ "task_categories:text-to-image", "annotations_creators:machine-generated", "language_creators:other", "multilinguality:monolingual", "size_categories:n<2K", "language:en", "license:cc-by-nc-sa-4.0", "region:us" ]

[]

[ "en" ]

[ "# Dataset Card for captioned Gundam\n\nScraped from URL (URL and manually cleaned to only keep drawings and \"Mobile Suits\" (i.e, humanoid-looking machines).\n\nThe captions were automatically generated from a generic hardcoded description + the dominant colors as described by BLIP." ]

[ "TAGS\n#task_categories-text-to-image #annotations_creators-machine-generated #language_creators-other #multilinguality-monolingual #size_categories-n<2K #language-English #license-cc-by-nc-sa-4.0 #region-us \n", "# Dataset Card for captioned Gundam\n\nScraped from URL (URL and manually cleaned to only keep drawings and \"Mobile Suits\" (i.e, humanoid-looking machines).\n\nThe captions were automatically generated from a generic hardcoded description + the dominant colors as described by BLIP." ]

[ "task_categories:fill-mask", "annotations_creators:no-annotation", "language_creators:found", "multilinguality:monolingual", "language:ca", "license:cc-by-nc-4.0", "region:us" ]

[]

[ "ca" ]

[ "# Dataset Card for Racó Forums Corpus", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Point of Contact: langtech@URL", "### Dataset Summary\n\nThe Racó Forums Corpus is a 19-million-sentence corpus of Catalan user-generated text built from the forums of Racó Català.\n\nSince the existing available corpora in Catalan lacked conversational data, we searched for a major source of such data for Catalan, and we found Racó Català, a popular multitopic online forum. We obtained a database dump and we transformed all the threads so that we obtained documents that traversed all the existing paths from the root (initial comment) to the leaves (last comment with no reply). In other words, if T is a tree such that T = {A,B,C,D} and the first comment is A that is replied by B and C independently, and, then, C is replied by D, we obtain two different documents A,B and A,C,D in the fairseq language modeling format.\n\nThis work is licensed under a Creative Commons Attribution Non-commercial 4.0 International License.", "### Supported Tasks and Leaderboards\n\nThis corpus is mainly intended to pretrain language models and word representations.", "### Languages\n\nThe dataset is in Catalan ('ca-ES').", "## Dataset Structure\n\nThe sentences are ordered to preserve the forum structure of comments and answers. T is a tree such that T = {A,B,C,D} and the first comment is A that is replied by B and C independently, and, then, C is replied by D, we obtain two different documents A,B and A,C,D in the fairseq language modeling format.", "### Data Instances", "### Data Splits\n\nThe dataset contains two splits: 'train' and 'valid'.", "## Dataset Creation", "### Curation Rationale\n\nWe created this corpus to contribute to the development of language models in Catalan, a low-resource language. The data was structured to preserve the dialogue structure of forums.", "### Source Data", "#### Initial Data Collection and Normalization\n\nThe data was structured and anonymized by the BSC.", "#### Who are the source language producers?\n\nThe data was provided by Racó Català.", "### Annotations\n\nThe dataset is unannotated.", "#### Annotation process\n\n[N/A]", "#### Who are the annotators?\n\n[N/A]", "### Personal and Sensitive Information\n\nThe data was annonymised to remove user names and emails, which were changed to random Catalan names. The mentions to the chat itself have also been changed.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nWe hope this corpus contributes to the development of language models in Catalan, a low-resource language.", "### Discussion of Biases\n\nWe are aware that, since the data comes from user-generated forums, this will contain biases, hate speech and toxic content. We have not applied any steps to reduce their impact.", "### Other Known Limitations\n\n[N/A]", "## Additional Information", "### Dataset Curators\n\nText Mining Unit (TeMU) at the Barcelona Supercomputing Center (bsc-temu@URL).\n\nThis work was funded by the Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya within the framework of Projecte AINA.", "### Licensing Information\n\nThis work is licensed under a Creative Commons Attribution Non-commercial 4.0 International License.", "### Contributions\n\nThanks to Racó Català for sharing their data." ]

[ "TAGS\n#task_categories-fill-mask #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #language-Catalan #license-cc-by-nc-4.0 #region-us \n", "# Dataset Card for Racó Forums Corpus", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Point of Contact: langtech@URL", "### Dataset Summary\n\nThe Racó Forums Corpus is a 19-million-sentence corpus of Catalan user-generated text built from the forums of Racó Català.\n\nSince the existing available corpora in Catalan lacked conversational data, we searched for a major source of such data for Catalan, and we found Racó Català, a popular multitopic online forum. We obtained a database dump and we transformed all the threads so that we obtained documents that traversed all the existing paths from the root (initial comment) to the leaves (last comment with no reply). In other words, if T is a tree such that T = {A,B,C,D} and the first comment is A that is replied by B and C independently, and, then, C is replied by D, we obtain two different documents A,B and A,C,D in the fairseq language modeling format.\n\nThis work is licensed under a Creative Commons Attribution Non-commercial 4.0 International License.", "### Supported Tasks and Leaderboards\n\nThis corpus is mainly intended to pretrain language models and word representations.", "### Languages\n\nThe dataset is in Catalan ('ca-ES').", "## Dataset Structure\n\nThe sentences are ordered to preserve the forum structure of comments and answers. T is a tree such that T = {A,B,C,D} and the first comment is A that is replied by B and C independently, and, then, C is replied by D, we obtain two different documents A,B and A,C,D in the fairseq language modeling format.", "### Data Instances", "### Data Splits\n\nThe dataset contains two splits: 'train' and 'valid'.", "## Dataset Creation", "### Curation Rationale\n\nWe created this corpus to contribute to the development of language models in Catalan, a low-resource language. The data was structured to preserve the dialogue structure of forums.", "### Source Data", "#### Initial Data Collection and Normalization\n\nThe data was structured and anonymized by the BSC.", "#### Who are the source language producers?\n\nThe data was provided by Racó Català.", "### Annotations\n\nThe dataset is unannotated.", "#### Annotation process\n\n[N/A]", "#### Who are the annotators?\n\n[N/A]", "### Personal and Sensitive Information\n\nThe data was annonymised to remove user names and emails, which were changed to random Catalan names. The mentions to the chat itself have also been changed.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nWe hope this corpus contributes to the development of language models in Catalan, a low-resource language.", "### Discussion of Biases\n\nWe are aware that, since the data comes from user-generated forums, this will contain biases, hate speech and toxic content. We have not applied any steps to reduce their impact.", "### Other Known Limitations\n\n[N/A]", "## Additional Information", "### Dataset Curators\n\nText Mining Unit (TeMU) at the Barcelona Supercomputing Center (bsc-temu@URL).\n\nThis work was funded by the Departament de la Vicepresidència i de Polítiques Digitals i Territori de la Generalitat de Catalunya within the framework of Projecte AINA.", "### Licensing Information\n\nThis work is licensed under a Creative Commons Attribution Non-commercial 4.0 International License.", "### Contributions\n\nThanks to Racó Català for sharing their data." ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"snli_shortcut_grammar\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"snli_shortcut_grammar\"\n\nMore Information needed" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"subj_shortcut_grammar\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"subj_shortcut_grammar\"\n\nMore Information needed" ]

[ "task_categories:text-classification", "task_categories:sentence-similarity", "task_ids:semantic-similarity-scoring", "annotations_creators:machine-generated", "language_creators:crowdsourced", "multilinguality:monolingual", "size_categories:10M<n<100M", "source_datasets:original", "language:en", "license:cc-by-sa-3.0", "wikipedia", "self-similarity", "arxiv:2012.03619", "region:us" ]

[ "2012.03619" ]

[ "en" ]

[ "# Dataset Card for 'wiki-paragraphs'", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information", "## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: URL\n- Leaderboard: \n- Point of Contact: Dennis Aumiller", "### Dataset Summary\n\nThe wiki-paragraphs dataset is constructed by automatically sampling two paragraphs from a Wikipedia article. If they are from the same section, they will be considered a \"semantic match\", otherwise as \"dissimilar\". Dissimilar paragraphs can in theory also be sampled from other documents, but have not shown any improvement in the particular evaluation of the linked work. \nThe alignment is in no way meant as an accurate depiction of similarity, but allows to quickly mine large amounts of samples.", "### Supported Tasks and Leaderboards\n\nThe dataset can be used for \"same-section classification\", which is a binary classification task (either two sentences/paragraphs belong to the same section or not).\nThis can be combined with document-level coherency measures, where we can check how many misclassifications appear within a single document.\nPlease refer to our paper for more details.", "### Languages\n\nThe data was extracted from English Wikipedia, therefore predominantly in English.", "## Dataset Structure", "### Data Instances\n\nA single instance contains three attributes:", "### Data Fields\n\n- sentence1: String containing the first paragraph\n- sentence2: String containing the second paragraph\n- label: Integer, either 0 or 1. Indicates whether two paragraphs belong to the same section (1) or come from different sections (0)", "### Data Splits\n\nWe provide train, validation and test splits, which were split as 80/10/10 from a randomly shuffled original data source.\nIn total, we provide 25375583 training pairs, as well as 3163685 validation and test instances, respectively.", "## Dataset Creation", "### Curation Rationale\n\nThe original idea was applied to self-segmentation of Terms of Service documents. Given that these are of domain-specific nature, we wanted to provide a more generally applicable model trained on Wikipedia data. \nIt is meant as a cheap-to-acquire pre-training strategy for large-scale experimentation with semantic similarity for long texts (paragraph-level).\nBased on our experiments, it is not necessarily sufficient by itself to replace traditional hand-labeled semantic similarity datasets.", "### Source Data", "#### Initial Data Collection and Normalization\n\nThe data was collected based on the articles considered in the Wiki-727k dataset by Koshorek et al. The dump of their dataset can be found through the respective Github repository. Note that we did *not* use the pre-processed data, but rather only information on the considered articles, which were re-acquired from Wikipedia at a more recent state.\nThis is due to the fact that paragraph information was not retained by the original Wiki-727k authors.\nWe did not verify the particular focus of considered pages.", "#### Who are the source language producers?\n\nWe do not have any further information on the contributors; these are volunteers contributing to URL.", "### Annotations", "#### Annotation process\n\nNo manual annotation was added to the dataset.\nWe automatically sampled two sections from within the same article; if these belong to the same section, they were assigned a label indicating the \"similarity\" (1), otherwise the label indicates that they are not belonging to the same section (0).\nWe sample three positive and three negative samples per section, per article.", "#### Who are the annotators?\n\nNo annotators were involved in the process.", "### Personal and Sensitive Information\n\nWe did not modify the original Wikipedia text in any way. Given that personal information, such as dates of birth (e.g., for a person of interest) may be on Wikipedia, this information is also considered in our dataset.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nThe purpose of the dataset is to serve as a *pre-training addition* for semantic similarity learning.\n\nSystems building on this dataset should consider additional, manually annotated data, before using a system in production.", "### Discussion of Biases\n\nTo our knowledge, there are some works indicating that male people have a several times larger chance of having a Wikipedia page created (especially in historical contexts). Therefore, a slight bias towards over-representation might be left in this dataset.", "### Other Known Limitations\n\nAs previously stated, the automatically extracted semantic similarity is not perfect; it should be treated as such.", "## Additional Information", "### Dataset Curators\n\nThe dataset was originally developed as a practical project by Lucienne-Sophie Marm� under the supervision of Dennis Aumiller.\nContributions to the original sampling strategy were made by Satya Almasian and Michael Gertz", "### Licensing Information\n\nWikipedia data is available under the CC-BY-SA 3.0 license." ]

[ "TAGS\n#task_categories-text-classification #task_categories-sentence-similarity #task_ids-semantic-similarity-scoring #annotations_creators-machine-generated #language_creators-crowdsourced #multilinguality-monolingual #size_categories-10M<n<100M #source_datasets-original #language-English #license-cc-by-sa-3.0 #wikipedia #self-similarity #arxiv-2012.03619 #region-us \n", "# Dataset Card for 'wiki-paragraphs'", "## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information", "## Dataset Description\n\n- Homepage: \n- Repository: URL\n- Paper: URL\n- Leaderboard: \n- Point of Contact: Dennis Aumiller", "### Dataset Summary\n\nThe wiki-paragraphs dataset is constructed by automatically sampling two paragraphs from a Wikipedia article. If they are from the same section, they will be considered a \"semantic match\", otherwise as \"dissimilar\". Dissimilar paragraphs can in theory also be sampled from other documents, but have not shown any improvement in the particular evaluation of the linked work. \nThe alignment is in no way meant as an accurate depiction of similarity, but allows to quickly mine large amounts of samples.", "### Supported Tasks and Leaderboards\n\nThe dataset can be used for \"same-section classification\", which is a binary classification task (either two sentences/paragraphs belong to the same section or not).\nThis can be combined with document-level coherency measures, where we can check how many misclassifications appear within a single document.\nPlease refer to our paper for more details.", "### Languages\n\nThe data was extracted from English Wikipedia, therefore predominantly in English.", "## Dataset Structure", "### Data Instances\n\nA single instance contains three attributes:", "### Data Fields\n\n- sentence1: String containing the first paragraph\n- sentence2: String containing the second paragraph\n- label: Integer, either 0 or 1. Indicates whether two paragraphs belong to the same section (1) or come from different sections (0)", "### Data Splits\n\nWe provide train, validation and test splits, which were split as 80/10/10 from a randomly shuffled original data source.\nIn total, we provide 25375583 training pairs, as well as 3163685 validation and test instances, respectively.", "## Dataset Creation", "### Curation Rationale\n\nThe original idea was applied to self-segmentation of Terms of Service documents. Given that these are of domain-specific nature, we wanted to provide a more generally applicable model trained on Wikipedia data. \nIt is meant as a cheap-to-acquire pre-training strategy for large-scale experimentation with semantic similarity for long texts (paragraph-level).\nBased on our experiments, it is not necessarily sufficient by itself to replace traditional hand-labeled semantic similarity datasets.", "### Source Data", "#### Initial Data Collection and Normalization\n\nThe data was collected based on the articles considered in the Wiki-727k dataset by Koshorek et al. The dump of their dataset can be found through the respective Github repository. Note that we did *not* use the pre-processed data, but rather only information on the considered articles, which were re-acquired from Wikipedia at a more recent state.\nThis is due to the fact that paragraph information was not retained by the original Wiki-727k authors.\nWe did not verify the particular focus of considered pages.", "#### Who are the source language producers?\n\nWe do not have any further information on the contributors; these are volunteers contributing to URL.", "### Annotations", "#### Annotation process\n\nNo manual annotation was added to the dataset.\nWe automatically sampled two sections from within the same article; if these belong to the same section, they were assigned a label indicating the \"similarity\" (1), otherwise the label indicates that they are not belonging to the same section (0).\nWe sample three positive and three negative samples per section, per article.", "#### Who are the annotators?\n\nNo annotators were involved in the process.", "### Personal and Sensitive Information\n\nWe did not modify the original Wikipedia text in any way. Given that personal information, such as dates of birth (e.g., for a person of interest) may be on Wikipedia, this information is also considered in our dataset.", "## Considerations for Using the Data", "### Social Impact of Dataset\n\nThe purpose of the dataset is to serve as a *pre-training addition* for semantic similarity learning.\n\nSystems building on this dataset should consider additional, manually annotated data, before using a system in production.", "### Discussion of Biases\n\nTo our knowledge, there are some works indicating that male people have a several times larger chance of having a Wikipedia page created (especially in historical contexts). Therefore, a slight bias towards over-representation might be left in this dataset.", "### Other Known Limitations\n\nAs previously stated, the automatically extracted semantic similarity is not perfect; it should be treated as such.", "## Additional Information", "### Dataset Curators\n\nThe dataset was originally developed as a practical project by Lucienne-Sophie Marm� under the supervision of Dennis Aumiller.\nContributions to the original sampling strategy were made by Satya Almasian and Michael Gertz", "### Licensing Information\n\nWikipedia data is available under the CC-BY-SA 3.0 license." ]

[ "benchmark:ttt", "region:us" ]

[]

[]

[ "# Batch job\n\nmodel_id: {model_id}\ndataset_name: {job.dataset_name}\ndataset_config: {job.dataset_config}\ndataset_split: {job.dataset_split}\ndataset_column: {job.dataset_column}" ]

[ "TAGS\n#benchmark-ttt #region-us \n", "# Batch job\n\nmodel_id: {model_id}\ndataset_name: {job.dataset_name}\ndataset_config: {job.dataset_config}\ndataset_split: {job.dataset_split}\ndataset_column: {job.dataset_column}" ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-560m\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: match\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: all\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b1\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-3b\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-1b7\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "autotrain", "evaluation", "region:us" ]

[]

[]

[ "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "TAGS\n#autotrain #evaluation #region-us \n", "# Dataset Card for AutoTrain Evaluator\n\nThis repository contains model predictions generated by AutoTrain for the following task and dataset:\n\n* Task: Zero-Shot Text Classification\n* Model: bigscience/bloom-7b1\n* Dataset: phpthinh/examplei\n* Config: mismatch\n* Split: test\n\nTo run new evaluation jobs, visit Hugging Face's automatic model evaluator.", "## Contributions\n\nThanks to @phpthinh for evaluating this model." ]

[ "task_categories:text2text-generation", "annotations_creators:found", "language_creators:found", "multilinguality:monolingual", "size_categories:100k<n<1M", "source_datasets:original", "language:bn", "license:cc-by-nc-sa-4.0", "conditional-text-generation", "paraphrase-generation", "arxiv:2210.05109", "region:us" ]

[ "2210.05109" ]

[ "bn" ]

[ "### Dataset Summary\n\n\nWe present BanglaParaphrase, a high quality synthetic Bangla paraphrase dataset containing about 466k paraphrase pairs.\nThe paraphrases ensures high quality by being semantically coherent and syntactically diverse.", "### Supported Tasks and Leaderboards\n\n\nMore information needed", "### Languages\n\n\n* 'bengali'\n\n\nLoading the dataset\n-------------------\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nOne example from the 'train' part of the dataset is given below in JSON format.", "### Data Fields\n\n\n* 'source': A string representing the source sentence.\n* 'target': A string representing the target sentence.", "### Data Splits\n\n\nDataset with train-dev-test example counts are given below:\n\n\n\nDataset Creation\n----------------", "### Curation Rationale\n\n\nMore information needed", "### Source Data\n\n\nRoar Bangla", "#### Initial Data Collection and Normalization\n\n\nDetailed in the paper", "#### Who are the source language producers?\n\n\nDetailed in the paper", "### Annotations\n\n\nDetailed in the paper", "#### Annotation process\n\n\nDetailed in the paper", "#### Who are the annotators?\n\n\nDetailed in the paper", "### Personal and Sensitive Information\n\n\nMore information needed\n\n\nConsiderations for Using the Data\n---------------------------------", "### Social Impact of Dataset\n\n\nMore information needed", "### Discussion of Biases\n\n\nMore information needed", "### Other Known Limitations\n\n\nMore information needed\n\n\nAdditional Information\n----------------------", "### Dataset Curators\n\n\nMore information needed", "### Licensing Information\n\n\nContents of this repository are restricted to only non-commercial research purposes under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0). Copyright of the dataset contents belongs to the original copyright holders.", "### Contributions" ]

[ "TAGS\n#task_categories-text2text-generation #annotations_creators-found #language_creators-found #multilinguality-monolingual #size_categories-100k<n<1M #source_datasets-original #language-Bengali #license-cc-by-nc-sa-4.0 #conditional-text-generation #paraphrase-generation #arxiv-2210.05109 #region-us \n", "### Dataset Summary\n\n\nWe present BanglaParaphrase, a high quality synthetic Bangla paraphrase dataset containing about 466k paraphrase pairs.\nThe paraphrases ensures high quality by being semantically coherent and syntactically diverse.", "### Supported Tasks and Leaderboards\n\n\nMore information needed", "### Languages\n\n\n* 'bengali'\n\n\nLoading the dataset\n-------------------\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nOne example from the 'train' part of the dataset is given below in JSON format.", "### Data Fields\n\n\n* 'source': A string representing the source sentence.\n* 'target': A string representing the target sentence.", "### Data Splits\n\n\nDataset with train-dev-test example counts are given below:\n\n\n\nDataset Creation\n----------------", "### Curation Rationale\n\n\nMore information needed", "### Source Data\n\n\nRoar Bangla", "#### Initial Data Collection and Normalization\n\n\nDetailed in the paper", "#### Who are the source language producers?\n\n\nDetailed in the paper", "### Annotations\n\n\nDetailed in the paper", "#### Annotation process\n\n\nDetailed in the paper", "#### Who are the annotators?\n\n\nDetailed in the paper", "### Personal and Sensitive Information\n\n\nMore information needed\n\n\nConsiderations for Using the Data\n---------------------------------", "### Social Impact of Dataset\n\n\nMore information needed", "### Discussion of Biases\n\n\nMore information needed", "### Other Known Limitations\n\n\nMore information needed\n\n\nAdditional Information\n----------------------", "### Dataset Curators\n\n\nMore information needed", "### Licensing Information\n\n\nContents of this repository are restricted to only non-commercial research purposes under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0). Copyright of the dataset contents belongs to the original copyright holders.", "### Contributions" ]

[ "region:us" ]

[]

[]

[]

[ "TAGS\n#region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"sidewalk-imagery\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"sidewalk-imagery\"\n\nMore Information needed" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "task_categories:conversational", "task_categories:question-answering", "task_categories:text-retrieval", "task_categories:visual-question-answering", "task_ids:open-domain-qa", "task_ids:extractive-qa", "task_ids:document-retrieval", "task_ids:visual-question-answering", "annotations_creators:no-annotation", "language_creators:found", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:afl-3.0", "youtube", "technical", "speech to text", "speech", "video", "video search", "audio", "audio search", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-conversational #task_categories-question-answering #task_categories-text-retrieval #task_categories-visual-question-answering #task_ids-open-domain-qa #task_ids-extractive-qa #task_ids-document-retrieval #task_ids-visual-question-answering #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-afl-3.0 #youtube #technical #speech to text #speech #video #video search #audio #audio search #region-us \n" ]

[ "task_categories:text-classification", "task_ids:natural-language-inference", "annotations_creators:expert-generated", "language_creators:expert-generated", "multilinguality:monolingual", "size_categories:n<1K", "source_datasets:original", "language:en", "license:cc-by-sa-4.0", "negation", "region:us" ]

[]

[ "en" ]

[ "# Dataset Card for [Dataset Name]", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage:\n- Repository:\n- Paper:\n- Leaderboard:\n- Point of Contact:", "### Dataset Summary", "### Supported Tasks and Leaderboards\n\nNatural Language Inference\nText Classification", "### Languages\n\nen", "## Dataset Structure", "### Data Instances", "### Data Fields\n\npremise:\nhypothesis:\nlabel:", "### Data Splits\n\nEvaluation: 258 samples", "## Dataset Creation", "### Curation Rationale\n\nExtracting samples corresponding to different linguistics constructions of negation.", "### Source Data\n\nGeoffrey K. Pullum and Rodney Huddleston. 2002. Negation, chapter 9. Cambridge University Press.", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?\n\nThe annotators are the authors of the papers, one of whom holds a graduate degree in linguistics.", "### Personal and Sensitive Information", "## Considerations for Using the Data", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information", "### Contributions\n\nThanks to @joey234 for adding this dataset." ]

[ "TAGS\n#task_categories-text-classification #task_ids-natural-language-inference #annotations_creators-expert-generated #language_creators-expert-generated #multilinguality-monolingual #size_categories-n<1K #source_datasets-original #language-English #license-cc-by-sa-4.0 #negation #region-us \n", "# Dataset Card for [Dataset Name]", "## Table of Contents\n- Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks and Leaderboards\n - Languages\n- Dataset Structure\n - Data Instances\n - Data Fields\n - Data Splits\n- Dataset Creation\n - Curation Rationale\n - Source Data\n - Annotations\n - Personal and Sensitive Information\n- Considerations for Using the Data\n - Social Impact of Dataset\n - Discussion of Biases\n - Other Known Limitations\n- Additional Information\n - Dataset Curators\n - Licensing Information\n - Citation Information\n - Contributions", "## Dataset Description\n\n- Homepage:\n- Repository:\n- Paper:\n- Leaderboard:\n- Point of Contact:", "### Dataset Summary", "### Supported Tasks and Leaderboards\n\nNatural Language Inference\nText Classification", "### Languages\n\nen", "## Dataset Structure", "### Data Instances", "### Data Fields\n\npremise:\nhypothesis:\nlabel:", "### Data Splits\n\nEvaluation: 258 samples", "## Dataset Creation", "### Curation Rationale\n\nExtracting samples corresponding to different linguistics constructions of negation.", "### Source Data\n\nGeoffrey K. Pullum and Rodney Huddleston. 2002. Negation, chapter 9. Cambridge University Press.", "#### Initial Data Collection and Normalization", "#### Who are the source language producers?", "### Annotations", "#### Annotation process", "#### Who are the annotators?\n\nThe annotators are the authors of the papers, one of whom holds a graduate degree in linguistics.", "### Personal and Sensitive Information", "## Considerations for Using the Data", "### Social Impact of Dataset", "### Discussion of Biases", "### Other Known Limitations", "## Additional Information", "### Dataset Curators", "### Licensing Information", "### Contributions\n\nThanks to @joey234 for adding this dataset." ]

[ "region:us" ]

[]

[]

[ "## test" ]

[ "TAGS\n#region-us \n", "## test" ]

[ "annotations_creators:found", "language_creators:found", "multilinguality:monolingual", "size_categories:100K<n<1M", "source_datasets:extended|wikipedia", "language:pt", "region:us" ]

[]

[ "pt" ]

[ "# Dataset Card for Dicionário Português\nIt is a list of 53138 portuguese words with its inflections.\n\n\nHow to use it: \n\nOutput:\n\nExemple: \n\nOutput:" ]

[ "TAGS\n#annotations_creators-found #language_creators-found #multilinguality-monolingual #size_categories-100K<n<1M #source_datasets-extended|wikipedia #language-Portuguese #region-us \n", "# Dataset Card for Dicionário Português\nIt is a list of 53138 portuguese words with its inflections.\n\n\nHow to use it: \n\nOutput:\n\nExemple: \n\nOutput:" ]

[ "task_categories:summarization", "task_categories:text2text-generation", "size_categories:1M<n<10M", "language:en", "license:apache-2.0", "conditional-text-generation", "region:us" ]

[]

[ "en" ]

[ "# About Dataset\n\nThe dataset consists of data from a bunch of youtube videos ranging from videos from fastai lessons, FSDL lesson to random videos teaching something.\nIn total this dataset contains 600 chapter markers in youtube and contains 25, 000 lesson transcript. \n\nThis dataset can be used for NLP tasks like summarization, topic segmentation etc. You can refer to some of the models we have trained with this dataset\nin github repo link for Full stack deep learning 2022 projects." ]

[ "TAGS\n#task_categories-summarization #task_categories-text2text-generation #size_categories-1M<n<10M #language-English #license-apache-2.0 #conditional-text-generation #region-us \n", "# About Dataset\n\nThe dataset consists of data from a bunch of youtube videos ranging from videos from fastai lessons, FSDL lesson to random videos teaching something.\nIn total this dataset contains 600 chapter markers in youtube and contains 25, 000 lesson transcript. \n\nThis dataset can be used for NLP tasks like summarization, topic segmentation etc. You can refer to some of the models we have trained with this dataset\nin github repo link for Full stack deep learning 2022 projects." ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"celeb-identities\"\n\nMore Information needed" ]

[ "task_categories:other", "annotations_creators:no-annotation", "language_creators:found", "multilinguality:monolingual", "size_categories:10K<n<100K", "source_datasets:original", "language:en", "license:afl-3.0", "youtube", "video", "video metadata", "tech", "science and tech", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-other #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-10K<n<100K #source_datasets-original #language-English #license-afl-3.0 #youtube #video #video metadata #tech #science and tech #region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"leaflet_offers-clone\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"leaflet_offers-clone\"\n\nMore Information needed" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"audio-diffusion-256-isolated-drums\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"audio-diffusion-256-isolated-drums\"\n\nMore Information needed" ]

[ "doi:10.57967/hf/0047", "region:us" ]

[]

[]

[]

[ "TAGS\n#doi-10.57967/hf/0047 #region-us \n" ]

[ "region:us" ]

[]

[]

[ "### Languages\n\n\nThe BCP-47 code for the dataset's language is unk.\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nA sample from this dataset looks as follows:", "### Dataset Fields\n\n\nThe dataset has the following fields (also called \"features\"):", "### Dataset Splits\n\n\nThis dataset is split into a train and validation split. The split sizes are as follow:" ]

[ "TAGS\n#region-us \n", "### Languages\n\n\nThe BCP-47 code for the dataset's language is unk.\n\n\nDataset Structure\n-----------------", "### Data Instances\n\n\nA sample from this dataset looks as follows:", "### Dataset Fields\n\n\nThe dataset has the following fields (also called \"features\"):", "### Dataset Splits\n\n\nThis dataset is split into a train and validation split. The split sizes are as follow:" ]

[ "task_categories:fill-mask", "multilinguality:monolingual", "size_categories:10K<n<100K", "language:en", "license:mit", "cloze", "mid-school", "high-school", "exams", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-fill-mask #multilinguality-monolingual #size_categories-10K<n<100K #language-English #license-mit #cloze #mid-school #high-school #exams #region-us \n" ]

[ "region:us" ]

[]

[]

[ "# wine-ratings\nProcessing, EDA, and ML on wine ratings" ]

[ "TAGS\n#region-us \n", "# wine-ratings\nProcessing, EDA, and ML on wine ratings" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"dummy_swin_pipe_5k\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"dummy_swin_pipe_5k\"\n\nMore Information needed" ]

[ "task_categories:fill-mask", "multilinguality:monolingual", "size_categories:1K<n<10K", "language:en", "license:mit", "cloze", "sciq", "mcql", "ai2 science questions", "region:us" ]

[]

[ "en" ]

[]

[ "TAGS\n#task_categories-fill-mask #multilinguality-monolingual #size_categories-1K<n<10K #language-English #license-mit #cloze #sciq #mcql #ai2 science questions #region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"figaro_hair_segmentation_1000\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"figaro_hair_segmentation_1000\"\n\nMore Information needed" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"dummy_swin_pipe\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"dummy_swin_pipe\"\n\nMore Information needed" ]

[ "region:us" ]

[]

[]

[]

[ "TAGS\n#region-us \n" ]

[ "license:cc-by-sa-4.0", "region:us" ]

[]

[]

[ "## Message Content Rephrasing Dataset\nIntroduced by Einolghozati et al. in Sound Natural: Content Rephrasing in Dialog Systems URL\n\nWe introduce a new task of rephrasing for amore natural virtual assistant. Currently, vir-tual assistants work in the paradigm of intent-slot tagging and the slot values are directlypassed as-is to the execution engine. However,this setup fails in some scenarios such as mes-saging when the query given by the user needsto be changed before repeating it or sending itto another user. For example, for queries like‘ask my wife if she can pick up the kids’ or ‘re-mind me to take my pills’, we need to rephrasethe content to ‘can you pick up the kids’ and‘take your pills’. In this paper, we study theproblem of rephrasing with messaging as ause case and release a dataset of 3000 pairs oforiginal query and rephrased query. We showthat BART, a pre-trained transformers-basedmasked language model with auto-regressivedecoding, is a strong baseline for the task, andshow improvements by adding a copy-pointerand copy loss to it. We analyze different trade-offs of BART-based and LSTM-based seq2seqmodels, and propose a distilled LSTM-basedseq2seq as the best practical model." ]

[ "TAGS\n#license-cc-by-sa-4.0 #region-us \n", "## Message Content Rephrasing Dataset\nIntroduced by Einolghozati et al. in Sound Natural: Content Rephrasing in Dialog Systems URL\n\nWe introduce a new task of rephrasing for amore natural virtual assistant. Currently, vir-tual assistants work in the paradigm of intent-slot tagging and the slot values are directlypassed as-is to the execution engine. However,this setup fails in some scenarios such as mes-saging when the query given by the user needsto be changed before repeating it or sending itto another user. For example, for queries like‘ask my wife if she can pick up the kids’ or ‘re-mind me to take my pills’, we need to rephrasethe content to ‘can you pick up the kids’ and‘take your pills’. In this paper, we study theproblem of rephrasing with messaging as ause case and release a dataset of 3000 pairs oforiginal query and rephrased query. We showthat BART, a pre-trained transformers-basedmasked language model with auto-regressivedecoding, is a strong baseline for the task, andshow improvements by adding a copy-pointerand copy loss to it. We analyze different trade-offs of BART-based and LSTM-based seq2seqmodels, and propose a distilled LSTM-basedseq2seq as the best practical model." ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"Quran_Hadith\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"Quran_Hadith\"\n\nMore Information needed" ]

[ "arxiv:2210.12478", "region:us" ]

[ "2210.12478" ]

[]

[]

[ "TAGS\n#arxiv-2210.12478 #region-us \n" ]

[ "region:us" ]

[]

[]

[ "# Dataset Card for \"AlRiyadh_Newspaper_Covid\"\n\nMore Information needed" ]

[ "TAGS\n#region-us \n", "# Dataset Card for \"AlRiyadh_Newspaper_Covid\"\n\nMore Information needed" ]