sha
stringlengths 40
40
| text
stringlengths 1
13.4M
| id
stringlengths 2
117
| tags
listlengths 1
7.91k
| created_at
stringlengths 25
25
| metadata
stringlengths 2
875k
| last_modified
stringlengths 25
25
| arxiv
listlengths 0
25
| languages
listlengths 0
7.91k
| tags_str
stringlengths 17
159k
| text_str
stringlengths 1
447k
| text_lists
listlengths 0
352
| processed_texts
listlengths 1
353
| tokens_length
listlengths 1
353
| input_texts
listlengths 1
40
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
40c19fb2b6e9f54fcdf0ad87c6266b1957b3b11b | # Dataset Card for "autotrain-data-3s6z-irs5-wkvn"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | healthcorum/autotrain-data-3s6z-irs5-wkvn | [
"region:us"
] | 2023-11-30T14:40:44+00:00 | {"dataset_info": {"features": [{"name": "autotrain_text", "dtype": "string"}, {"name": "autotrain_label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 28804645, "num_examples": 7998}, {"name": "validation", "num_bytes": 7203538, "num_examples": 2000}], "download_size": 5982807, "dataset_size": 36008183}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}]}]} | 2023-11-30T14:40:46+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "autotrain-data-3s6z-irs5-wkvn"
More Information needed | [
"# Dataset Card for \"autotrain-data-3s6z-irs5-wkvn\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"autotrain-data-3s6z-irs5-wkvn\"\n\nMore Information needed"
] | [
6,
26
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"autotrain-data-3s6z-irs5-wkvn\"\n\nMore Information needed"
] |
f7a6939ca8fedcc89f35c513f10321ea5994f738 | # Dataset Card for MuST-SHE_en-ca
## Dataset Description
### Dataset Summary
MuST-SHE_en-ca is an English-Catalan evaluation dataset of **1.046** samples, created to support evaluation of Catalan NLP tasks,
e.g., Gender Bias detection in Machine Translation. This dataset is derived from MuST-SHE English-Spanish, created using a combination of automatic and human translation.
For more information about the original MuST-SHE dataset, the structure of which has been preserved, refer to: https://mt.fbk.eu/must-she/
### Supported Tasks
The dataset can be used to evaluate Gender Bias in Machine Translation from English to Catalan.
### Languages
It supports translation from English to Catalan.
## Dataset Structure
One tsv file with 1.046 rows.
- MuST-SHE_en-ca.tsv
Overall it follows the same structure as the original MuST-SHE dataset. Nevertheless, an extra column named "TEXT-CATEGORY" has been added.
This column is meant to adapt the dataset for Machine Translation (text). In this scenario, all instances are divided into two categories: with and
without contextual information to disambiguate gender, instead of the original four categories.
### Data fields
Following the original structure in original MuST-SHE.
- ID
- LANG
- TALK
- MuSTC-v1.0-SET
- SRC
- REF
- WRONG-REF
- SPEAKER
- GENDER
- CATEGORY
- TEXT-CATEGORY
- COMMENT
- FREE-REF
- GENDERTERMS
### Data Splits
The dataset contains a single split for evaluation.
## Source Data
The original MuST-SHE is a subset of the TED-based MuST-C corpus.
MuST-SHE_en-ca was created by automatically translating the Spanish components of the English-Spanish MuST-SHE using the [PlanTL Project's Spanish-Catalan machine translation model](https://huggingface.co/PlanTL-GOB-ES/mt-plantl-es-ca). Gender terms were extracted automatically and both the gender terms and automatically translated sentences were then extensively revised by a native Catalan speaker to ensure accuracy.
### Personal and Sensitive Information
No anonymisation process was performed.
## Considerations for Using the Data
### Social Impact of Dataset
The purpose of this dataset is to help evaluating the inherent bias of Machines Translation engines for low-resource languages such as Catalan.
## Additional Information
### Author
Language Technologies Unit (LangTech) at the Barcelona Supercomputing Center.
### Contact information
For further information, please send an email to <[email protected]>.
### Copyright
Copyright Language Technologies Unit at Barcelona Supercomputing Center (2023).
### Licensing information
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 (CC BY NC ND 4.0 International) license.
### Funding
This work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (https://projecteaina.cat/). | projecte-aina/MuST-SHE_en-ca | [
"license:cc",
"region:us"
] | 2023-11-30T14:49:51+00:00 | {"license": "cc"} | 2024-01-17T13:53:30+00:00 | [] | [] | TAGS
#license-cc #region-us
| # Dataset Card for MuST-SHE_en-ca
## Dataset Description
### Dataset Summary
MuST-SHE_en-ca is an English-Catalan evaluation dataset of 1.046 samples, created to support evaluation of Catalan NLP tasks,
e.g., Gender Bias detection in Machine Translation. This dataset is derived from MuST-SHE English-Spanish, created using a combination of automatic and human translation.
For more information about the original MuST-SHE dataset, the structure of which has been preserved, refer to: URL
### Supported Tasks
The dataset can be used to evaluate Gender Bias in Machine Translation from English to Catalan.
### Languages
It supports translation from English to Catalan.
## Dataset Structure
One tsv file with 1.046 rows.
- MuST-SHE_en-URL
Overall it follows the same structure as the original MuST-SHE dataset. Nevertheless, an extra column named "TEXT-CATEGORY" has been added.
This column is meant to adapt the dataset for Machine Translation (text). In this scenario, all instances are divided into two categories: with and
without contextual information to disambiguate gender, instead of the original four categories.
### Data fields
Following the original structure in original MuST-SHE.
- ID
- LANG
- TALK
- MuSTC-v1.0-SET
- SRC
- REF
- WRONG-REF
- SPEAKER
- GENDER
- CATEGORY
- TEXT-CATEGORY
- COMMENT
- FREE-REF
- GENDERTERMS
### Data Splits
The dataset contains a single split for evaluation.
## Source Data
The original MuST-SHE is a subset of the TED-based MuST-C corpus.
MuST-SHE_en-ca was created by automatically translating the Spanish components of the English-Spanish MuST-SHE using the PlanTL Project's Spanish-Catalan machine translation model. Gender terms were extracted automatically and both the gender terms and automatically translated sentences were then extensively revised by a native Catalan speaker to ensure accuracy.
### Personal and Sensitive Information
No anonymisation process was performed.
## Considerations for Using the Data
### Social Impact of Dataset
The purpose of this dataset is to help evaluating the inherent bias of Machines Translation engines for low-resource languages such as Catalan.
## Additional Information
### Author
Language Technologies Unit (LangTech) at the Barcelona Supercomputing Center.
### Contact information
For further information, please send an email to <langtech@URL>.
### Copyright
Copyright Language Technologies Unit at Barcelona Supercomputing Center (2023).
### Licensing information
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 (CC BY NC ND 4.0 International) license.
### Funding
This work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL | [
"# Dataset Card for MuST-SHE_en-ca",
"## Dataset Description",
"### Dataset Summary\n\nMuST-SHE_en-ca is an English-Catalan evaluation dataset of 1.046 samples, created to support evaluation of Catalan NLP tasks,\ne.g., Gender Bias detection in Machine Translation. This dataset is derived from MuST-SHE English-Spanish, created using a combination of automatic and human translation. \n\nFor more information about the original MuST-SHE dataset, the structure of which has been preserved, refer to: URL",
"### Supported Tasks\n\nThe dataset can be used to evaluate Gender Bias in Machine Translation from English to Catalan.",
"### Languages\n\nIt supports translation from English to Catalan.",
"## Dataset Structure\n\nOne tsv file with 1.046 rows.\n\n- MuST-SHE_en-URL\n\nOverall it follows the same structure as the original MuST-SHE dataset. Nevertheless, an extra column named \"TEXT-CATEGORY\" has been added.\nThis column is meant to adapt the dataset for Machine Translation (text). In this scenario, all instances are divided into two categories: with and\nwithout contextual information to disambiguate gender, instead of the original four categories.",
"### Data fields\nFollowing the original structure in original MuST-SHE.\n - ID\n - LANG\n - TALK\n - MuSTC-v1.0-SET\n - SRC\n - REF\n - WRONG-REF\n - SPEAKER\n - GENDER\n - CATEGORY\n - TEXT-CATEGORY\n - COMMENT\n - FREE-REF\n - GENDERTERMS",
"### Data Splits\n\nThe dataset contains a single split for evaluation.",
"## Source Data\n\nThe original MuST-SHE is a subset of the TED-based MuST-C corpus.\nMuST-SHE_en-ca was created by automatically translating the Spanish components of the English-Spanish MuST-SHE using the PlanTL Project's Spanish-Catalan machine translation model. Gender terms were extracted automatically and both the gender terms and automatically translated sentences were then extensively revised by a native Catalan speaker to ensure accuracy.",
"### Personal and Sensitive Information\n\nNo anonymisation process was performed.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe purpose of this dataset is to help evaluating the inherent bias of Machines Translation engines for low-resource languages such as Catalan.",
"## Additional Information",
"### Author\nLanguage Technologies Unit (LangTech) at the Barcelona Supercomputing Center.",
"### Contact information\nFor further information, please send an email to <langtech@URL>.",
"### Copyright\nCopyright Language Technologies Unit at Barcelona Supercomputing Center (2023).",
"### Licensing information\nThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 (CC BY NC ND 4.0 International) license.",
"### Funding\nThis work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL"
] | [
"TAGS\n#license-cc #region-us \n",
"# Dataset Card for MuST-SHE_en-ca",
"## Dataset Description",
"### Dataset Summary\n\nMuST-SHE_en-ca is an English-Catalan evaluation dataset of 1.046 samples, created to support evaluation of Catalan NLP tasks,\ne.g., Gender Bias detection in Machine Translation. This dataset is derived from MuST-SHE English-Spanish, created using a combination of automatic and human translation. \n\nFor more information about the original MuST-SHE dataset, the structure of which has been preserved, refer to: URL",
"### Supported Tasks\n\nThe dataset can be used to evaluate Gender Bias in Machine Translation from English to Catalan.",
"### Languages\n\nIt supports translation from English to Catalan.",
"## Dataset Structure\n\nOne tsv file with 1.046 rows.\n\n- MuST-SHE_en-URL\n\nOverall it follows the same structure as the original MuST-SHE dataset. Nevertheless, an extra column named \"TEXT-CATEGORY\" has been added.\nThis column is meant to adapt the dataset for Machine Translation (text). In this scenario, all instances are divided into two categories: with and\nwithout contextual information to disambiguate gender, instead of the original four categories.",
"### Data fields\nFollowing the original structure in original MuST-SHE.\n - ID\n - LANG\n - TALK\n - MuSTC-v1.0-SET\n - SRC\n - REF\n - WRONG-REF\n - SPEAKER\n - GENDER\n - CATEGORY\n - TEXT-CATEGORY\n - COMMENT\n - FREE-REF\n - GENDERTERMS",
"### Data Splits\n\nThe dataset contains a single split for evaluation.",
"## Source Data\n\nThe original MuST-SHE is a subset of the TED-based MuST-C corpus.\nMuST-SHE_en-ca was created by automatically translating the Spanish components of the English-Spanish MuST-SHE using the PlanTL Project's Spanish-Catalan machine translation model. Gender terms were extracted automatically and both the gender terms and automatically translated sentences were then extensively revised by a native Catalan speaker to ensure accuracy.",
"### Personal and Sensitive Information\n\nNo anonymisation process was performed.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe purpose of this dataset is to help evaluating the inherent bias of Machines Translation engines for low-resource languages such as Catalan.",
"## Additional Information",
"### Author\nLanguage Technologies Unit (LangTech) at the Barcelona Supercomputing Center.",
"### Contact information\nFor further information, please send an email to <langtech@URL>.",
"### Copyright\nCopyright Language Technologies Unit at Barcelona Supercomputing Center (2023).",
"### Licensing information\nThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 (CC BY NC ND 4.0 International) license.",
"### Funding\nThis work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL"
] | [
11,
14,
4,
111,
27,
13,
122,
79,
16,
111,
16,
8,
39,
5,
21,
20,
18,
37,
27
] | [
"passage: TAGS\n#license-cc #region-us \n# Dataset Card for MuST-SHE_en-ca## Dataset Description### Dataset Summary\n\nMuST-SHE_en-ca is an English-Catalan evaluation dataset of 1.046 samples, created to support evaluation of Catalan NLP tasks,\ne.g., Gender Bias detection in Machine Translation. This dataset is derived from MuST-SHE English-Spanish, created using a combination of automatic and human translation. \n\nFor more information about the original MuST-SHE dataset, the structure of which has been preserved, refer to: URL### Supported Tasks\n\nThe dataset can be used to evaluate Gender Bias in Machine Translation from English to Catalan.### Languages\n\nIt supports translation from English to Catalan.## Dataset Structure\n\nOne tsv file with 1.046 rows.\n\n- MuST-SHE_en-URL\n\nOverall it follows the same structure as the original MuST-SHE dataset. Nevertheless, an extra column named \"TEXT-CATEGORY\" has been added.\nThis column is meant to adapt the dataset for Machine Translation (text). In this scenario, all instances are divided into two categories: with and\nwithout contextual information to disambiguate gender, instead of the original four categories.### Data fields\nFollowing the original structure in original MuST-SHE.\n - ID\n - LANG\n - TALK\n - MuSTC-v1.0-SET\n - SRC\n - REF\n - WRONG-REF\n - SPEAKER\n - GENDER\n - CATEGORY\n - TEXT-CATEGORY\n - COMMENT\n - FREE-REF\n - GENDERTERMS### Data Splits\n\nThe dataset contains a single split for evaluation."
] |
ad41ed8dce723a9dab4b6e495b1a3a4ab8fd04c9 | # Dataset Card for "processed_t5_small_context_len_512"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | yardeny/processed_t5_small_context_len_512 | [
"region:us"
] | 2023-11-30T14:50:37+00:00 | {"dataset_info": {"features": [{"name": "input_ids", "sequence": "int32"}, {"name": "attention_mask", "sequence": "int8"}], "splits": [{"name": "train", "num_bytes": 17763456912.0, "num_examples": 6917234}], "download_size": 6975491955, "dataset_size": 17763456912.0}} | 2023-11-30T15:02:31+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "processed_t5_small_context_len_512"
More Information needed | [
"# Dataset Card for \"processed_t5_small_context_len_512\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"processed_t5_small_context_len_512\"\n\nMore Information needed"
] | [
6,
26
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"processed_t5_small_context_len_512\"\n\nMore Information needed"
] |
f19a64dbe45a354c0399d98ac181ccfbb747cbed |
# Dataset Card for CA-EN Parallel Corpus
## Table of Contents
- [Dataset Description](#dataset-description)
- [Dataset Summary](#dataset-summary)
- [Supported Tasks](#supported-tasks-and-leaderboards)
- [Languages](#languages)
- [Dataset Structure](#dataset-structure)
- [Data Splits](#data-instances)
- [Dataset Creation](#dataset-creation)
- [Source Data](#source-data)
- [Data preparation](#data-preparation)
- [Personal and Sensitive Information](#personal-and-sensitive-information)
- [Considerations for Using the Data](#considerations-for-using-the-data)
- [Social Impact of Dataset](#social-impact-of-dataset)
- [Discussion of Biases](#discussion-of-biases)
- [Other Known Limitations](#other-known-limitations)
- [Additional Information](#additional-information)
- [Author](#author)
- [Contact Information](#contact-information)
- [Copyright](#copyright)
- [Licensing information](#licenciung-informatrion)
- [Funding](#funding)
## Dataset Description
### Dataset Summary
The CA-EN Parallel Corpus is a Catalan-English dataset of **14.385.296** parallel sentences. The dataset was created to support Catalan NLP tasks, e.g.,
Machine Translation.
### Supported Tasks and Leaderboards
The dataset can be used to train a model for Multilingual Machine Translation. Success on this task is typically measured by achieving a high BLEU score.
### Languages
The texts in the dataset are in Catalan and English.
## Dataset Structure
The dataset is a single tsv file where each row contains a paralle sentence pair.
### Data Splits
The dataset contains a single split: `train`.
## Dataset Creation
### Source Data
The The data is a brand new collection of parallel sentences in Catalan and English partially derived from web crawlings and belonging to a mix of different domains and styles.
The source data is Catalan authentic text translated to English or authentic English text translated to Catalan.
### Data preparation
The data was obtained through a combination of human translation and machine translation with human proofreading.
After the translation process, the data was deduplicated and filtered to remove any sentence pairs with a cosine similarity of less than 0.75 in order to improve the data alignment quality.
This was done using sentence embeddings calculated using [LaBSE](https://huggingface.co/sentence-transformers/LaBSE).
The obtained cleaned corpus consists of **14.385.296** parallel sentences of human quality.
### Personal and Sensitive Information
No anonymisation process was performed.
## Considerations for Using the Data
### Social Impact of Dataset
The purpose of this dataset is to help develop Machine Translation tasks for mid-resource languages such as Catalan.
### Discussion of Biases
We are aware that since part of the data comes from unreliable web pages and non-curated texts, some biases may be present in the dataset.
Nonetheless, we have not applied any steps to reduce their impact.
### Other Known Limitations
The dataset contains data of a general domain. Application of this dataset in more specific domains such as biomedical, legal etc. would be of limited use.
## Additional Information
### Author
Language Technologies Unit (LangTech) at the Barcelona Supercomputing Center.
### Contact information
For further information, please send an email to [email protected].
### Copyright
Copyright Language Technologies Unit at Barcelona Supercomputing Center (2023).
### Licensing information
This work is licensed under a [Creative Commons Attribution 4.0 International licence](https://creativecommons.org/licenses/by/4.0/).
### Funding
This work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (https://projecteaina.cat/). | projecte-aina/CA-EN_Parallel_Corpus | [
"task_categories:translation",
"multilinguality:translation",
"size_categories:10M<n<100M",
"source_datasets:original",
"language:ca",
"language:en",
"language:multilingual",
"license:cc-by-4.0",
"region:us"
] | 2023-11-30T14:52:47+00:00 | {"language": ["ca", "en", "multilingual"], "license": "cc-by-4.0", "multilinguality": ["translation"], "size_categories": ["10M<n<100M"], "source_datasets": ["original"], "task_categories": ["translation"], "task_ids": [], "pretty_name": "CA-EN Parallel Corpus"} | 2024-01-17T13:52:58+00:00 | [] | [
"ca",
"en",
"multilingual"
] | TAGS
#task_categories-translation #multilinguality-translation #size_categories-10M<n<100M #source_datasets-original #language-Catalan #language-English #language-multilingual #license-cc-by-4.0 #region-us
|
# Dataset Card for CA-EN Parallel Corpus
## Table of Contents
- Dataset Description
- Dataset Summary
- Supported Tasks
- Languages
- Dataset Structure
- Data Splits
- Dataset Creation
- Source Data
- Data preparation
- Personal and Sensitive Information
- Considerations for Using the Data
- Social Impact of Dataset
- Discussion of Biases
- Other Known Limitations
- Additional Information
- Author
- Contact Information
- Copyright
- Licensing information
- Funding
## Dataset Description
### Dataset Summary
The CA-EN Parallel Corpus is a Catalan-English dataset of 14.385.296 parallel sentences. The dataset was created to support Catalan NLP tasks, e.g.,
Machine Translation.
### Supported Tasks and Leaderboards
The dataset can be used to train a model for Multilingual Machine Translation. Success on this task is typically measured by achieving a high BLEU score.
### Languages
The texts in the dataset are in Catalan and English.
## Dataset Structure
The dataset is a single tsv file where each row contains a paralle sentence pair.
### Data Splits
The dataset contains a single split: 'train'.
## Dataset Creation
### Source Data
The The data is a brand new collection of parallel sentences in Catalan and English partially derived from web crawlings and belonging to a mix of different domains and styles.
The source data is Catalan authentic text translated to English or authentic English text translated to Catalan.
### Data preparation
The data was obtained through a combination of human translation and machine translation with human proofreading.
After the translation process, the data was deduplicated and filtered to remove any sentence pairs with a cosine similarity of less than 0.75 in order to improve the data alignment quality.
This was done using sentence embeddings calculated using LaBSE.
The obtained cleaned corpus consists of 14.385.296 parallel sentences of human quality.
### Personal and Sensitive Information
No anonymisation process was performed.
## Considerations for Using the Data
### Social Impact of Dataset
The purpose of this dataset is to help develop Machine Translation tasks for mid-resource languages such as Catalan.
### Discussion of Biases
We are aware that since part of the data comes from unreliable web pages and non-curated texts, some biases may be present in the dataset.
Nonetheless, we have not applied any steps to reduce their impact.
### Other Known Limitations
The dataset contains data of a general domain. Application of this dataset in more specific domains such as biomedical, legal etc. would be of limited use.
## Additional Information
### Author
Language Technologies Unit (LangTech) at the Barcelona Supercomputing Center.
### Contact information
For further information, please send an email to langtech@URL.
### Copyright
Copyright Language Technologies Unit at Barcelona Supercomputing Center (2023).
### Licensing information
This work is licensed under a Creative Commons Attribution 4.0 International licence.
### Funding
This work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL | [
"# Dataset Card for CA-EN Parallel Corpus",
"## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks\n - Languages\n- Dataset Structure\n - Data Splits\n- Dataset Creation\n - Source Data\n - Data preparation\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 - Author\n - Contact Information\n - Copyright\n - Licensing information\n - Funding",
"## Dataset Description",
"### Dataset Summary\n\nThe CA-EN Parallel Corpus is a Catalan-English dataset of 14.385.296 parallel sentences. The dataset was created to support Catalan NLP tasks, e.g., \nMachine Translation.",
"### Supported Tasks and Leaderboards\n\nThe dataset can be used to train a model for Multilingual Machine Translation. Success on this task is typically measured by achieving a high BLEU score.",
"### Languages\n\nThe texts in the dataset are in Catalan and English.",
"## Dataset Structure\n\nThe dataset is a single tsv file where each row contains a paralle sentence pair.",
"### Data Splits\n\nThe dataset contains a single split: 'train'.",
"## Dataset Creation",
"### Source Data\n\nThe The data is a brand new collection of parallel sentences in Catalan and English partially derived from web crawlings and belonging to a mix of different domains and styles.\nThe source data is Catalan authentic text translated to English or authentic English text translated to Catalan.",
"### Data preparation\n\n The data was obtained through a combination of human translation and machine translation with human proofreading.\n After the translation process, the data was deduplicated and filtered to remove any sentence pairs with a cosine similarity of less than 0.75 in order to improve the data alignment quality.\n This was done using sentence embeddings calculated using LaBSE.\n The obtained cleaned corpus consists of 14.385.296 parallel sentences of human quality.",
"### Personal and Sensitive Information\n\nNo anonymisation process was performed.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe purpose of this dataset is to help develop Machine Translation tasks for mid-resource languages such as Catalan.",
"### Discussion of Biases\n\nWe are aware that since part of the data comes from unreliable web pages and non-curated texts, some biases may be present in the dataset.\nNonetheless, we have not applied any steps to reduce their impact.",
"### Other Known Limitations\n\nThe dataset contains data of a general domain. Application of this dataset in more specific domains such as biomedical, legal etc. would be of limited use.",
"## Additional Information",
"### Author\nLanguage Technologies Unit (LangTech) at the Barcelona Supercomputing Center.",
"### Contact information\nFor further information, please send an email to langtech@URL.",
"### Copyright\nCopyright Language Technologies Unit at Barcelona Supercomputing Center (2023).",
"### Licensing information\nThis work is licensed under a Creative Commons Attribution 4.0 International licence.",
"### Funding\nThis work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL"
] | [
"TAGS\n#task_categories-translation #multilinguality-translation #size_categories-10M<n<100M #source_datasets-original #language-Catalan #language-English #language-multilingual #license-cc-by-4.0 #region-us \n",
"# Dataset Card for CA-EN Parallel Corpus",
"## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks\n - Languages\n- Dataset Structure\n - Data Splits\n- Dataset Creation\n - Source Data\n - Data preparation\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 - Author\n - Contact Information\n - Copyright\n - Licensing information\n - Funding",
"## Dataset Description",
"### Dataset Summary\n\nThe CA-EN Parallel Corpus is a Catalan-English dataset of 14.385.296 parallel sentences. The dataset was created to support Catalan NLP tasks, e.g., \nMachine Translation.",
"### Supported Tasks and Leaderboards\n\nThe dataset can be used to train a model for Multilingual Machine Translation. Success on this task is typically measured by achieving a high BLEU score.",
"### Languages\n\nThe texts in the dataset are in Catalan and English.",
"## Dataset Structure\n\nThe dataset is a single tsv file where each row contains a paralle sentence pair.",
"### Data Splits\n\nThe dataset contains a single split: 'train'.",
"## Dataset Creation",
"### Source Data\n\nThe The data is a brand new collection of parallel sentences in Catalan and English partially derived from web crawlings and belonging to a mix of different domains and styles.\nThe source data is Catalan authentic text translated to English or authentic English text translated to Catalan.",
"### Data preparation\n\n The data was obtained through a combination of human translation and machine translation with human proofreading.\n After the translation process, the data was deduplicated and filtered to remove any sentence pairs with a cosine similarity of less than 0.75 in order to improve the data alignment quality.\n This was done using sentence embeddings calculated using LaBSE.\n The obtained cleaned corpus consists of 14.385.296 parallel sentences of human quality.",
"### Personal and Sensitive Information\n\nNo anonymisation process was performed.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe purpose of this dataset is to help develop Machine Translation tasks for mid-resource languages such as Catalan.",
"### Discussion of Biases\n\nWe are aware that since part of the data comes from unreliable web pages and non-curated texts, some biases may be present in the dataset.\nNonetheless, we have not applied any steps to reduce their impact.",
"### Other Known Limitations\n\nThe dataset contains data of a general domain. Application of this dataset in more specific domains such as biomedical, legal etc. would be of limited use.",
"## Additional Information",
"### Author\nLanguage Technologies Unit (LangTech) at the Barcelona Supercomputing Center.",
"### Contact information\nFor further information, please send an email to langtech@URL.",
"### Copyright\nCopyright Language Technologies Unit at Barcelona Supercomputing Center (2023).",
"### Licensing information\nThis work is licensed under a Creative Commons Attribution 4.0 International licence.",
"### Funding\nThis work has been promoted and financed by the Generalitat de Catalunya through the [Aina project] (URL"
] | [
66,
10,
97,
4,
49,
45,
17,
27,
19,
5,
64,
105,
16,
8,
32,
61,
43,
5,
21,
18,
18,
20,
27
] | [
"passage: TAGS\n#task_categories-translation #multilinguality-translation #size_categories-10M<n<100M #source_datasets-original #language-Catalan #language-English #language-multilingual #license-cc-by-4.0 #region-us \n# Dataset Card for CA-EN Parallel Corpus## Table of Contents\n- Dataset Description\n - Dataset Summary\n - Supported Tasks\n - Languages\n- Dataset Structure\n - Data Splits\n- Dataset Creation\n - Source Data\n - Data preparation\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 - Author\n - Contact Information\n - Copyright\n - Licensing information\n - Funding## Dataset Description### Dataset Summary\n\nThe CA-EN Parallel Corpus is a Catalan-English dataset of 14.385.296 parallel sentences. The dataset was created to support Catalan NLP tasks, e.g., \nMachine Translation.### Supported Tasks and Leaderboards\n\nThe dataset can be used to train a model for Multilingual Machine Translation. Success on this task is typically measured by achieving a high BLEU score.### Languages\n\nThe texts in the dataset are in Catalan and English.## Dataset Structure\n\nThe dataset is a single tsv file where each row contains a paralle sentence pair.### Data Splits\n\nThe dataset contains a single split: 'train'.## Dataset Creation### Source Data\n\nThe The data is a brand new collection of parallel sentences in Catalan and English partially derived from web crawlings and belonging to a mix of different domains and styles.\nThe source data is Catalan authentic text translated to English or authentic English text translated to Catalan."
] |
f40167f165b59d624dbe9afb050b665256336b3a | # Trelis Function Calling Dataset - VERSION 3
Access this dataset by purchasing a license [HERE](https://buy.stripe.com/eVa28U6xTeqfcV25li).
- Allows models to be fine-tuned for function-calling.
- The dataset is human generated and does not make use of Llama 2 or OpenAI!
- The dataset includes 66 training rows, 19 validation rows and 5 test rows (for manual evaluation).
- Based on eight functions: search_bing, search_arxiv, save_chat, read_json_file, list_files, get_current_weather, delete_file, clear_chat
Alternatively, you can find pre-trained function calling models on [Trelis Mart](https://mart.trelis.com)
## Updates since v2
- Cross-compatible function format: The format now matches OpenAI's function format, making it easy to migrate from using OpenAI APIs to any models fine-tuned with this dataset.
- Chain function calling: Ability (particularly with larger models) to first make a call to one function in order to get data for a second function call.
- Supported by inferencing scripts, read more below.
--Change-log--
04Dec2023 - Official release of function_calling_v3
02Dec2023 - Pre-release of function_calling_v3
## Inference Scripts
Out-of-the-box inference scripts are available for purchase:
- Purchase only the function calling inference scripts, [HERE](https://buy.stripe.com/28o00M9K50zp4ow4hf)
- Purchase as part of the full ADVANCED-inference repo, [HERE](https://trelis.com/enterprise-server-api-and-inference-guide/).
## Fine-Tuning Notes and Scripts
The objective of function calling is for the model to return a structured json object *and nothing else*. The performance of fine-tuning depends **strongly** on how the attention mask and loss mask are set. For further details see the [Youtube Video Here](https://youtu.be/OQdp-OeG1as).
The fine-tuning script is available for purchase alone [here](https://buy.stripe.com/fZe14Qe0l81R9IQaFy), or is included in the ADVANCED-fine-tuning repository available for purchase on [Trelis.com](https://trelis.com).
### QLoRa Training Notebook for Llama 2 (FREE)
- Access a basic Google Colab script for fine-tuning [here](https://colab.research.google.com/drive/1uMSS1o_8YOPyG1X_4k6ENEE3kJfBGGhH?usp=sharing).
## Licensing
The Function Calling Extended dataset is suitable for commercial use.
Further terms:
- Licenses are not transferable to other users/entities.
- The dataset may not be re-published in it's current or derivative form.
- The dataset may be used to train and fine-tune commercial language models.
### Attribution of data sources
This project includes data from the TruthfulQA dataset, which is available at: https://huggingface.co/datasets/truthful_qa. The truthful_qa dataset is licensed under the Apache License 2.0, Copyright (C) 2023, Stephanie Lin, Jacob Hilton, and Owain Evans.
## Prompt Format (example below is for openchat)
```
B_FUNC, E_FUNC = "You have access to the following functions. Use them if required:\n\n", "\n\n"
B_INST, E_INST = "GPT4 Correct User: ", "<|end_of_turn|>GPT4 Correct Assistant:" #OpenChat style
# B_INST, E_INST = "[INST] ", " [/INST]" #Llama 2 style
functionList = data['test'][index]['functionList']
user_prompt = data['test'][index]['userPrompt']
correct_answer = data['test'][index]['assistantResponse']
prompt = f"{E_FUNC}{B_FUNC}{functionList.strip()}{E_FUNC}{B_INST}{user_prompt.strip()}{E_INST}\n\n"
```
## Sample Prompt and Response:
```
You have access to the following functions. Use them if required:
[
{
"type": "function",
"function": {
"name": "get_stock_price",
"description": "Get the stock price of an array of stocks",
"parameters": {
"type": "object",
"properties": {
"names": {
"type": "array",
"items": {
"type": "string"
},
"description": "An array of stocks"
}
},
"required": [
"names"
]
}
}
},
{
"type": "function",
"function": {
"name": "get_big_stocks",
"description": "Get the names of the largest N stocks by market cap",
"parameters": {
"type": "object",
"properties": {
"number": {
"type": "integer",
"description": "The number of largest stocks to get the names of, e.g. 25"
},
"region": {
"type": "string",
"description": "The region to consider, can be \"US\" or \"World\"."
}
},
"required": [
"number"
]
}
}
}
]GPT4 Correct User: Get the price of Apple's stock<|end_of_turn|>GPT4 Correct Assistant:{
"name": "get_stock_price",
"arguments": {
"names": [
"Apple"
]
}
}<|end_of_turn|>
```
## CSV File Structure
The generated CSV file has the following columns:
- `functionList`: Descriptions of two functions (the current function and a randomly selected other function).
- `userPrompt`: The user's prompt.
- `assistantResponse`: The assistant's response.
### JSON File Structure
Function metadata format follows the OpenAI standard.
Each function file should be a JSON file with the following structure:
```json
{
"type": "function",
"function": {
"name": "function_name",
"description": "function description",
"parameters": {
"type": "object",
"properties": {
"property_1": {
"type": "property_type", //#e.g. string
"description": "property description"
},
"property_2": {
"type": "property_type", //#e.g. string
"description": "property description"
}
},
"required": ["property_1","property_2"]
}
},
"samplePromptResponsePairs": [
{
"prompt": "sample_prompt",
"response": {
"name": "generate_password",
"arguments": {
"property_1": "property_value",
"property_2": "property_value"
}
}
},
...
]
}
```
The `functionMetaData` object describes the function. The `samplePromptResponsePairs` array contains sample prompts and responses for the function.
### Testing JSON Structure
A script named `validate.py` can be used to validate the structure of a function JSON file. It checks for the presence and correct types of all necessary keys in the JSON structure.
To use the script, call it from the command line with the name of the function file as an argument:
```
python validate.py my_function.json
```
## Repo Structure (for prompt dataset generation)
- `functions/`: This directory contains function files, each of which is a JSON file with a specific structure that describes a function and its sample prompts and responses.
- `generate_dataset.py`: This Python script generates the base training and testing dataset CSV files. The first example in each function json file is used in the validation dataset and the rest are used for the train dataset.
- `addBlank.py`: This adds in truthfulqa questions and answers after system prompts with functions.
- `text_responses.py`: adds in prompts to accustomise the model to the presence of function descriptions at the start of prompt sequences.
There are also, some equivalent files for generating a test dataset - to be used for manual evaluation:
- `test_functions/`: contains functions for manual evaluation, different to the training and test set of functions.
- create_test_datasets.py - which runs createTestPrompts.py and test_text_responses.py
- createTestPrompts.py which creates data rows to test function calling without and without required arguments provided, as well as one chain function calling test (e.g. where one function must be called before the other).
- test_text_responses.py generates data rows to test out simple prompts (e.g. Greetings!), short non-sensical prompts (e.g. "shop"), and also a standard question (What planets are in our solar system?). | Trelis/function_calling_v3 | [
"task_categories:question-answering",
"task_categories:conversational",
"task_categories:text-generation",
"size_categories:n<1K",
"language:en",
"function call",
"function calling",
"function-calling",
"region:us"
] | 2023-11-30T15:10:18+00:00 | {"language": ["en"], "size_categories": ["n<1K"], "task_categories": ["question-answering", "conversational", "text-generation"], "tags": ["function call", "function calling", "function-calling"], "extra_gated_prompt": "Purchase access to this repo [HERE](https://buy.stripe.com/eVa28U6xTeqfcV25li)", "extra_gated_fields": {"I have purchased a license (access will be granted once your payment clears)": "checkbox", "I agree to the terms of the license described on the dataset card": "checkbox"}} | 2023-12-04T14:33:35+00:00 | [] | [
"en"
] | TAGS
#task_categories-question-answering #task_categories-conversational #task_categories-text-generation #size_categories-n<1K #language-English #function call #function calling #function-calling #region-us
| # Trelis Function Calling Dataset - VERSION 3
Access this dataset by purchasing a license HERE.
- Allows models to be fine-tuned for function-calling.
- The dataset is human generated and does not make use of Llama 2 or OpenAI!
- The dataset includes 66 training rows, 19 validation rows and 5 test rows (for manual evaluation).
- Based on eight functions: search_bing, search_arxiv, save_chat, read_json_file, list_files, get_current_weather, delete_file, clear_chat
Alternatively, you can find pre-trained function calling models on Trelis Mart
## Updates since v2
- Cross-compatible function format: The format now matches OpenAI's function format, making it easy to migrate from using OpenAI APIs to any models fine-tuned with this dataset.
- Chain function calling: Ability (particularly with larger models) to first make a call to one function in order to get data for a second function call.
- Supported by inferencing scripts, read more below.
--Change-log--
04Dec2023 - Official release of function_calling_v3
02Dec2023 - Pre-release of function_calling_v3
## Inference Scripts
Out-of-the-box inference scripts are available for purchase:
- Purchase only the function calling inference scripts, HERE
- Purchase as part of the full ADVANCED-inference repo, HERE.
## Fine-Tuning Notes and Scripts
The objective of function calling is for the model to return a structured json object *and nothing else*. The performance of fine-tuning depends strongly on how the attention mask and loss mask are set. For further details see the Youtube Video Here.
The fine-tuning script is available for purchase alone here, or is included in the ADVANCED-fine-tuning repository available for purchase on URL.
### QLoRa Training Notebook for Llama 2 (FREE)
- Access a basic Google Colab script for fine-tuning here.
## Licensing
The Function Calling Extended dataset is suitable for commercial use.
Further terms:
- Licenses are not transferable to other users/entities.
- The dataset may not be re-published in it's current or derivative form.
- The dataset may be used to train and fine-tune commercial language models.
### Attribution of data sources
This project includes data from the TruthfulQA dataset, which is available at: URL The truthful_qa dataset is licensed under the Apache License 2.0, Copyright (C) 2023, Stephanie Lin, Jacob Hilton, and Owain Evans.
## Prompt Format (example below is for openchat)
## Sample Prompt and Response:
## CSV File Structure
The generated CSV file has the following columns:
- 'functionList': Descriptions of two functions (the current function and a randomly selected other function).
- 'userPrompt': The user's prompt.
- 'assistantResponse': The assistant's response.
### JSON File Structure
Function metadata format follows the OpenAI standard.
Each function file should be a JSON file with the following structure:
The 'functionMetaData' object describes the function. The 'samplePromptResponsePairs' array contains sample prompts and responses for the function.
### Testing JSON Structure
A script named 'URL' can be used to validate the structure of a function JSON file. It checks for the presence and correct types of all necessary keys in the JSON structure.
To use the script, call it from the command line with the name of the function file as an argument:
## Repo Structure (for prompt dataset generation)
- 'functions/': This directory contains function files, each of which is a JSON file with a specific structure that describes a function and its sample prompts and responses.
- 'generate_dataset.py': This Python script generates the base training and testing dataset CSV files. The first example in each function json file is used in the validation dataset and the rest are used for the train dataset.
- 'URL': This adds in truthfulqa questions and answers after system prompts with functions.
- 'text_responses.py': adds in prompts to accustomise the model to the presence of function descriptions at the start of prompt sequences.
There are also, some equivalent files for generating a test dataset - to be used for manual evaluation:
- 'test_functions/': contains functions for manual evaluation, different to the training and test set of functions.
- create_test_datasets.py - which runs URL and test_text_responses.py
- URL which creates data rows to test function calling without and without required arguments provided, as well as one chain function calling test (e.g. where one function must be called before the other).
- test_text_responses.py generates data rows to test out simple prompts (e.g. Greetings!), short non-sensical prompts (e.g. "shop"), and also a standard question (What planets are in our solar system?). | [
"# Trelis Function Calling Dataset - VERSION 3\n\nAccess this dataset by purchasing a license HERE.\n- Allows models to be fine-tuned for function-calling.\n- The dataset is human generated and does not make use of Llama 2 or OpenAI!\n- The dataset includes 66 training rows, 19 validation rows and 5 test rows (for manual evaluation).\n- Based on eight functions: search_bing, search_arxiv, save_chat, read_json_file, list_files, get_current_weather, delete_file, clear_chat\n\nAlternatively, you can find pre-trained function calling models on Trelis Mart",
"## Updates since v2\n- Cross-compatible function format: The format now matches OpenAI's function format, making it easy to migrate from using OpenAI APIs to any models fine-tuned with this dataset.\n- Chain function calling: Ability (particularly with larger models) to first make a call to one function in order to get data for a second function call.\n- Supported by inferencing scripts, read more below.\n\n--Change-log--\n\n04Dec2023 - Official release of function_calling_v3\n\n02Dec2023 - Pre-release of function_calling_v3",
"## Inference Scripts\nOut-of-the-box inference scripts are available for purchase:\n- Purchase only the function calling inference scripts, HERE\n- Purchase as part of the full ADVANCED-inference repo, HERE.",
"## Fine-Tuning Notes and Scripts\n\nThe objective of function calling is for the model to return a structured json object *and nothing else*. The performance of fine-tuning depends strongly on how the attention mask and loss mask are set. For further details see the Youtube Video Here.\n\nThe fine-tuning script is available for purchase alone here, or is included in the ADVANCED-fine-tuning repository available for purchase on URL.",
"### QLoRa Training Notebook for Llama 2 (FREE)\n- Access a basic Google Colab script for fine-tuning here.",
"## Licensing\nThe Function Calling Extended dataset is suitable for commercial use.\n\nFurther terms:\n- Licenses are not transferable to other users/entities.\n- The dataset may not be re-published in it's current or derivative form.\n- The dataset may be used to train and fine-tune commercial language models.",
"### Attribution of data sources\n\nThis project includes data from the TruthfulQA dataset, which is available at: URL The truthful_qa dataset is licensed under the Apache License 2.0, Copyright (C) 2023, Stephanie Lin, Jacob Hilton, and Owain Evans.",
"## Prompt Format (example below is for openchat)",
"## Sample Prompt and Response:",
"## CSV File Structure\n\nThe generated CSV file has the following columns:\n\n- 'functionList': Descriptions of two functions (the current function and a randomly selected other function).\n- 'userPrompt': The user's prompt.\n- 'assistantResponse': The assistant's response.",
"### JSON File Structure\n\nFunction metadata format follows the OpenAI standard.\n\nEach function file should be a JSON file with the following structure:\n\n\n\nThe 'functionMetaData' object describes the function. The 'samplePromptResponsePairs' array contains sample prompts and responses for the function.",
"### Testing JSON Structure\n\nA script named 'URL' can be used to validate the structure of a function JSON file. It checks for the presence and correct types of all necessary keys in the JSON structure.\n\nTo use the script, call it from the command line with the name of the function file as an argument:",
"## Repo Structure (for prompt dataset generation)\n\n- 'functions/': This directory contains function files, each of which is a JSON file with a specific structure that describes a function and its sample prompts and responses.\n- 'generate_dataset.py': This Python script generates the base training and testing dataset CSV files. The first example in each function json file is used in the validation dataset and the rest are used for the train dataset.\n- 'URL': This adds in truthfulqa questions and answers after system prompts with functions.\n- 'text_responses.py': adds in prompts to accustomise the model to the presence of function descriptions at the start of prompt sequences.\n\nThere are also, some equivalent files for generating a test dataset - to be used for manual evaluation:\n- 'test_functions/': contains functions for manual evaluation, different to the training and test set of functions.\n- create_test_datasets.py - which runs URL and test_text_responses.py\n- URL which creates data rows to test function calling without and without required arguments provided, as well as one chain function calling test (e.g. where one function must be called before the other).\n- test_text_responses.py generates data rows to test out simple prompts (e.g. Greetings!), short non-sensical prompts (e.g. \"shop\"), and also a standard question (What planets are in our solar system?)."
] | [
"TAGS\n#task_categories-question-answering #task_categories-conversational #task_categories-text-generation #size_categories-n<1K #language-English #function call #function calling #function-calling #region-us \n",
"# Trelis Function Calling Dataset - VERSION 3\n\nAccess this dataset by purchasing a license HERE.\n- Allows models to be fine-tuned for function-calling.\n- The dataset is human generated and does not make use of Llama 2 or OpenAI!\n- The dataset includes 66 training rows, 19 validation rows and 5 test rows (for manual evaluation).\n- Based on eight functions: search_bing, search_arxiv, save_chat, read_json_file, list_files, get_current_weather, delete_file, clear_chat\n\nAlternatively, you can find pre-trained function calling models on Trelis Mart",
"## Updates since v2\n- Cross-compatible function format: The format now matches OpenAI's function format, making it easy to migrate from using OpenAI APIs to any models fine-tuned with this dataset.\n- Chain function calling: Ability (particularly with larger models) to first make a call to one function in order to get data for a second function call.\n- Supported by inferencing scripts, read more below.\n\n--Change-log--\n\n04Dec2023 - Official release of function_calling_v3\n\n02Dec2023 - Pre-release of function_calling_v3",
"## Inference Scripts\nOut-of-the-box inference scripts are available for purchase:\n- Purchase only the function calling inference scripts, HERE\n- Purchase as part of the full ADVANCED-inference repo, HERE.",
"## Fine-Tuning Notes and Scripts\n\nThe objective of function calling is for the model to return a structured json object *and nothing else*. The performance of fine-tuning depends strongly on how the attention mask and loss mask are set. For further details see the Youtube Video Here.\n\nThe fine-tuning script is available for purchase alone here, or is included in the ADVANCED-fine-tuning repository available for purchase on URL.",
"### QLoRa Training Notebook for Llama 2 (FREE)\n- Access a basic Google Colab script for fine-tuning here.",
"## Licensing\nThe Function Calling Extended dataset is suitable for commercial use.\n\nFurther terms:\n- Licenses are not transferable to other users/entities.\n- The dataset may not be re-published in it's current or derivative form.\n- The dataset may be used to train and fine-tune commercial language models.",
"### Attribution of data sources\n\nThis project includes data from the TruthfulQA dataset, which is available at: URL The truthful_qa dataset is licensed under the Apache License 2.0, Copyright (C) 2023, Stephanie Lin, Jacob Hilton, and Owain Evans.",
"## Prompt Format (example below is for openchat)",
"## Sample Prompt and Response:",
"## CSV File Structure\n\nThe generated CSV file has the following columns:\n\n- 'functionList': Descriptions of two functions (the current function and a randomly selected other function).\n- 'userPrompt': The user's prompt.\n- 'assistantResponse': The assistant's response.",
"### JSON File Structure\n\nFunction metadata format follows the OpenAI standard.\n\nEach function file should be a JSON file with the following structure:\n\n\n\nThe 'functionMetaData' object describes the function. The 'samplePromptResponsePairs' array contains sample prompts and responses for the function.",
"### Testing JSON Structure\n\nA script named 'URL' can be used to validate the structure of a function JSON file. It checks for the presence and correct types of all necessary keys in the JSON structure.\n\nTo use the script, call it from the command line with the name of the function file as an argument:",
"## Repo Structure (for prompt dataset generation)\n\n- 'functions/': This directory contains function files, each of which is a JSON file with a specific structure that describes a function and its sample prompts and responses.\n- 'generate_dataset.py': This Python script generates the base training and testing dataset CSV files. The first example in each function json file is used in the validation dataset and the rest are used for the train dataset.\n- 'URL': This adds in truthfulqa questions and answers after system prompts with functions.\n- 'text_responses.py': adds in prompts to accustomise the model to the presence of function descriptions at the start of prompt sequences.\n\nThere are also, some equivalent files for generating a test dataset - to be used for manual evaluation:\n- 'test_functions/': contains functions for manual evaluation, different to the training and test set of functions.\n- create_test_datasets.py - which runs URL and test_text_responses.py\n- URL which creates data rows to test function calling without and without required arguments provided, as well as one chain function calling test (e.g. where one function must be called before the other).\n- test_text_responses.py generates data rows to test out simple prompts (e.g. Greetings!), short non-sensical prompts (e.g. \"shop\"), and also a standard question (What planets are in our solar system?)."
] | [
64,
157,
138,
57,
100,
30,
76,
60,
15,
9,
72,
74,
73,
348
] | [
"passage: TAGS\n#task_categories-question-answering #task_categories-conversational #task_categories-text-generation #size_categories-n<1K #language-English #function call #function calling #function-calling #region-us \n# Trelis Function Calling Dataset - VERSION 3\n\nAccess this dataset by purchasing a license HERE.\n- Allows models to be fine-tuned for function-calling.\n- The dataset is human generated and does not make use of Llama 2 or OpenAI!\n- The dataset includes 66 training rows, 19 validation rows and 5 test rows (for manual evaluation).\n- Based on eight functions: search_bing, search_arxiv, save_chat, read_json_file, list_files, get_current_weather, delete_file, clear_chat\n\nAlternatively, you can find pre-trained function calling models on Trelis Mart## Updates since v2\n- Cross-compatible function format: The format now matches OpenAI's function format, making it easy to migrate from using OpenAI APIs to any models fine-tuned with this dataset.\n- Chain function calling: Ability (particularly with larger models) to first make a call to one function in order to get data for a second function call.\n- Supported by inferencing scripts, read more below.\n\n--Change-log--\n\n04Dec2023 - Official release of function_calling_v3\n\n02Dec2023 - Pre-release of function_calling_v3## Inference Scripts\nOut-of-the-box inference scripts are available for purchase:\n- Purchase only the function calling inference scripts, HERE\n- Purchase as part of the full ADVANCED-inference repo, HERE.",
"passage: ## Fine-Tuning Notes and Scripts\n\nThe objective of function calling is for the model to return a structured json object *and nothing else*. The performance of fine-tuning depends strongly on how the attention mask and loss mask are set. For further details see the Youtube Video Here.\n\nThe fine-tuning script is available for purchase alone here, or is included in the ADVANCED-fine-tuning repository available for purchase on URL.### QLoRa Training Notebook for Llama 2 (FREE)\n- Access a basic Google Colab script for fine-tuning here.## Licensing\nThe Function Calling Extended dataset is suitable for commercial use.\n\nFurther terms:\n- Licenses are not transferable to other users/entities.\n- The dataset may not be re-published in it's current or derivative form.\n- The dataset may be used to train and fine-tune commercial language models.### Attribution of data sources\n\nThis project includes data from the TruthfulQA dataset, which is available at: URL The truthful_qa dataset is licensed under the Apache License 2.0, Copyright (C) 2023, Stephanie Lin, Jacob Hilton, and Owain Evans.## Prompt Format (example below is for openchat)## Sample Prompt and Response:## CSV File Structure\n\nThe generated CSV file has the following columns:\n\n- 'functionList': Descriptions of two functions (the current function and a randomly selected other function).\n- 'userPrompt': The user's prompt.\n- 'assistantResponse': The assistant's response.### JSON File Structure\n\nFunction metadata format follows the OpenAI standard.\n\nEach function file should be a JSON file with the following structure:\n\n\n\nThe 'functionMetaData' object describes the function. The 'samplePromptResponsePairs' array contains sample prompts and responses for the function.### Testing JSON Structure\n\nA script named 'URL' can be used to validate the structure of a function JSON file. It checks for the presence and correct types of all necessary keys in the JSON structure.\n\nTo use the script, call it from the command line with the name of the function file as an argument:"
] |
432df097c29af7199ed5653fbd0739edf1c3acbc |
ARBRES-Kenstur
==============
ARBRES-Kenstur is a Breton-French parallel corpora generated by extracting the French translations of Breton sentences from the interlinear [glosses](https://en.wikipedia.org/wiki/Interlinear_gloss) of the [ARBRES](https://arbres.iker.cnrs.fr) wikigrammar.
The extraction is still under developpment in the [Autogramm project](https://autogramm.github.io/en/) of the French National Research Agency. More information can be found on their [Github repository](https://github.com/Autogramm/Breton). | lgrobol/ARBRES-Kenstur | [
"task_categories:translation",
"size_categories:1K<n<10K",
"language:br",
"language:fr",
"license:cc-by-sa-4.0",
"region:us"
] | 2023-11-30T15:42:59+00:00 | {"language": ["br", "fr"], "license": "cc-by-sa-4.0", "size_categories": ["1K<n<10K"], "task_categories": ["translation"]} | 2023-11-30T15:49:22+00:00 | [] | [
"br",
"fr"
] | TAGS
#task_categories-translation #size_categories-1K<n<10K #language-Breton #language-French #license-cc-by-sa-4.0 #region-us
|
ARBRES-Kenstur
==============
ARBRES-Kenstur is a Breton-French parallel corpora generated by extracting the French translations of Breton sentences from the interlinear glosses of the ARBRES wikigrammar.
The extraction is still under developpment in the Autogramm project of the French National Research Agency. More information can be found on their Github repository. | [] | [
"TAGS\n#task_categories-translation #size_categories-1K<n<10K #language-Breton #language-French #license-cc-by-sa-4.0 #region-us \n"
] | [
50
] | [
"passage: TAGS\n#task_categories-translation #size_categories-1K<n<10K #language-Breton #language-French #license-cc-by-sa-4.0 #region-us \n"
] |
4d3f0b666754a46551ed82c0cd33945151c25f25 | # Dataset Card for "mnli"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | veluribharath/mnli | [
"region:us"
] | 2023-11-30T15:43:56+00:00 | {"dataset_info": {"features": [{"name": "premise", "dtype": "string"}, {"name": "hypothesis", "dtype": "string"}, {"name": "label", "dtype": {"class_label": {"names": {"0": "entailment", "1": "neutral", "2": "contradiction"}}}}, {"name": "id", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 76190454, "num_examples": 392702}, {"name": "validation_matched", "num_bytes": 1873043, "num_examples": 9815}, {"name": "validation_mismatched", "num_bytes": 1988559, "num_examples": 9832}, {"name": "test_matched", "num_bytes": 1887838, "num_examples": 9796}, {"name": "test_mismatched", "num_bytes": 1990091, "num_examples": 9847}], "download_size": 57193931, "dataset_size": 83929985}} | 2023-11-30T15:48:23+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "mnli"
More Information needed | [
"# Dataset Card for \"mnli\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"mnli\"\n\nMore Information needed"
] | [
6,
12
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"mnli\"\n\nMore Information needed"
] |
d65a682068a5aaf2416b01d2bcd40946181262e0 | # Dataset Card for "AdjectiveScaleProbe-nli"
```bib
@inproceedings{10.1609/aaai.v37i11.26559,
author = {Liu, Wei and Xiang, Ming and Ding, Nai},
title = {Adjective Scale Probe: Can Language Models Encode Formal Semantics Information?},
year = {2023},
isbn = {978-1-57735-880-0},
publisher = {AAAI Press},
url = {https://doi.org/10.1609/aaai.v37i11.26559},
doi = {10.1609/aaai.v37i11.26559},
booktitle = {Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence and Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence},
articleno = {1490},
numpages = {9},
series = {AAAI'23/IAAI'23/EAAI'23}
}
``` | tasksource/AdjectiveScaleProbe-nli | [
"region:us"
] | 2023-11-30T15:57:22+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}, {"split": "test", "path": "data/test-*"}]}], "dataset_info": {"features": [{"name": "premise", "dtype": "string"}, {"name": "hypothesis", "dtype": "string"}, {"name": "label", "dtype": "string"}, {"name": "config", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1673129, "num_examples": 15600}, {"name": "validation", "num_bytes": 402347, "num_examples": 3748}, {"name": "test", "num_bytes": 283098, "num_examples": 3084}], "download_size": 0, "dataset_size": 2358574}} | 2023-12-05T10:23:12+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "AdjectiveScaleProbe-nli"
| [
"# Dataset Card for \"AdjectiveScaleProbe-nli\""
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"AdjectiveScaleProbe-nli\""
] | [
6,
17
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"AdjectiveScaleProbe-nli\""
] |
7717a57cfe209f334351de66c683d0b088fb715b | # Data Engineer: Take home project
## Introduction
The goal of this project is to evaluate your knowledge and skills in the design and implementation of a scalable data pre-processing pipeline.
## Problem statement
- Reads audio data being populated by Metavoice product, `Studio`, into a CloudFlare R2 bucket
- Runs two data transformation steps on the audio files:
- Transcription - use [Whisper](https://github.com/openai/whisper)
- Tokenisation - use mock code [here](https://gist.github.com/sidroopdaska/364e9f493d8dd9584eb9e1e9cae5715c)
- Stores the results using the example schema below.
```<id - relative path of audio file>, <transcription>, <token array>```
## Requirements
- Install `ffmpeg` by following instructions [here](https://www.hostinger.com/tutorials/how-to-install-ffmpeg)
- Use pipenv to install the required packages:
```pipenv install```
- Go to where the `main.py` file is located and run:
```python main.py ```
## Notes
For scalability, I decided to read the audio file with a given chunk_size, and so preprocess the audio file in chunks.
This is to avoid memory issues when dealing with large audio files.
The script is broken after a while (probably an audio file it does not like) as it shows:
```pydub.exceptions.CouldntDecodeError: Decoding failed. ffmpeg returned error code: 1```
I think there is a better solution, but by lack of time and not 100% sure if that feasible, that would be to:
- create a HuggingFace [loading-script](https://huggingface.co/docs/datasets/audio_dataset#loading-script)
- And so we could use the HF Dataset API to load the audio files and preprocess it.
- For the Whisper model, HF provide useful functions to [preprocess](https://huggingface.co/learn/audio-course/chapter1/preprocessing) it:
```
from transformers import WhisperFeatureExtractor
feature_extractor = WhisperFeatureExtractor.from_pretrained("openai/whisper-small")
...
```
| FlorianD/metavoice | [
"region:us"
] | 2023-11-30T16:16:10+00:00 | {} | 2023-12-06T15:36:38+00:00 | [] | [] | TAGS
#region-us
| # Data Engineer: Take home project
## Introduction
The goal of this project is to evaluate your knowledge and skills in the design and implementation of a scalable data pre-processing pipeline.
## Problem statement
- Reads audio data being populated by Metavoice product, 'Studio', into a CloudFlare R2 bucket
- Runs two data transformation steps on the audio files:
- Transcription - use Whisper
- Tokenisation - use mock code here
- Stores the results using the example schema below.
## Requirements
- Install 'ffmpeg' by following instructions here
- Use pipenv to install the required packages:
- Go to where the 'URL' file is located and run:
## Notes
For scalability, I decided to read the audio file with a given chunk_size, and so preprocess the audio file in chunks.
This is to avoid memory issues when dealing with large audio files.
The script is broken after a while (probably an audio file it does not like) as it shows:
I think there is a better solution, but by lack of time and not 100% sure if that feasible, that would be to:
- create a HuggingFace loading-script
- And so we could use the HF Dataset API to load the audio files and preprocess it.
- For the Whisper model, HF provide useful functions to preprocess it:
| [
"# Data Engineer: Take home project",
"## Introduction\nThe goal of this project is to evaluate your knowledge and skills in the design and implementation of a scalable data pre-processing pipeline.",
"## Problem statement\n- Reads audio data being populated by Metavoice product, 'Studio', into a CloudFlare R2 bucket\n- Runs two data transformation steps on the audio files:\n - Transcription - use Whisper\n - Tokenisation - use mock code here\n- Stores the results using the example schema below.",
"## Requirements\n- Install 'ffmpeg' by following instructions here\n- Use pipenv to install the required packages:\n \n- Go to where the 'URL' file is located and run:",
"## Notes\nFor scalability, I decided to read the audio file with a given chunk_size, and so preprocess the audio file in chunks. \nThis is to avoid memory issues when dealing with large audio files.\nThe script is broken after a while (probably an audio file it does not like) as it shows:\n\n\n\nI think there is a better solution, but by lack of time and not 100% sure if that feasible, that would be to:\n- create a HuggingFace loading-script\n- And so we could use the HF Dataset API to load the audio files and preprocess it.\n- For the Whisper model, HF provide useful functions to preprocess it:"
] | [
"TAGS\n#region-us \n",
"# Data Engineer: Take home project",
"## Introduction\nThe goal of this project is to evaluate your knowledge and skills in the design and implementation of a scalable data pre-processing pipeline.",
"## Problem statement\n- Reads audio data being populated by Metavoice product, 'Studio', into a CloudFlare R2 bucket\n- Runs two data transformation steps on the audio files:\n - Transcription - use Whisper\n - Tokenisation - use mock code here\n- Stores the results using the example schema below.",
"## Requirements\n- Install 'ffmpeg' by following instructions here\n- Use pipenv to install the required packages:\n \n- Go to where the 'URL' file is located and run:",
"## Notes\nFor scalability, I decided to read the audio file with a given chunk_size, and so preprocess the audio file in chunks. \nThis is to avoid memory issues when dealing with large audio files.\nThe script is broken after a while (probably an audio file it does not like) as it shows:\n\n\n\nI think there is a better solution, but by lack of time and not 100% sure if that feasible, that would be to:\n- create a HuggingFace loading-script\n- And so we could use the HF Dataset API to load the audio files and preprocess it.\n- For the Whisper model, HF provide useful functions to preprocess it:"
] | [
6,
8,
33,
72,
42,
148
] | [
"passage: TAGS\n#region-us \n# Data Engineer: Take home project## Introduction\nThe goal of this project is to evaluate your knowledge and skills in the design and implementation of a scalable data pre-processing pipeline.## Problem statement\n- Reads audio data being populated by Metavoice product, 'Studio', into a CloudFlare R2 bucket\n- Runs two data transformation steps on the audio files:\n - Transcription - use Whisper\n - Tokenisation - use mock code here\n- Stores the results using the example schema below.## Requirements\n- Install 'ffmpeg' by following instructions here\n- Use pipenv to install the required packages:\n \n- Go to where the 'URL' file is located and run:## Notes\nFor scalability, I decided to read the audio file with a given chunk_size, and so preprocess the audio file in chunks. \nThis is to avoid memory issues when dealing with large audio files.\nThe script is broken after a while (probably an audio file it does not like) as it shows:\n\n\n\nI think there is a better solution, but by lack of time and not 100% sure if that feasible, that would be to:\n- create a HuggingFace loading-script\n- And so we could use the HF Dataset API to load the audio files and preprocess it.\n- For the Whisper model, HF provide useful functions to preprocess it:"
] |
e70430b058de2a294b7e507aa27e4c0298708341 |
# Dataset Card for "DrugsCom Reviews"
## Table of Contents
- [Dataset Description](#dataset-description)
- [Dataset Summary](#dataset-summary)
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
- [Languages](#languages)
- [Dataset Structure](#dataset-structure)
- [Data Instances](#data-instances)
- [Data Fields](#data-fields)
- [Data Splits](#data-splits)
- [Dataset Creation](#dataset-creation)
- [Curation Rationale](#curation-rationale)
- [Source Data](#source-data)
- [Annotations](#annotations)
- [Considerations for Using the Data](#considerations-for-using-the-data)
- [Social Impact of Dataset](#social-impact-of-dataset)
- [Discussion of Biases](#discussion-of-biases)
- [Other Known Limitations](#other-known-limitations)
- [Additional Information](#additional-information)
- [Dataset Curators](#dataset-curators)
- [Licensing Information](#licensing-information)
- [Citation Information](#citation-information)
- [Contributions](#contributions)
## Dataset Description
- **Homepage:** [UCI Machine Learning Repository](https://archive.ics.uci.edu/dataset/462/drug+review+dataset+drugs+com)
- **Repository:** [DrugsCom Reviews on Hugging Face](https://huggingface.co/datasets/Zakia/drugscom_reviews)
- **Size of the training dataset file:** 82.314 MB
- **Size of the test dataset file:** 27.414 MB
- **Total size of downloaded dataset files:** 109.73 MB
### Dataset Summary
The DrugsCom Reviews dataset is originally sourced from the UCI Machine Learning Repository. It provides patient reviews on specific drugs along with related conditions and a 10-star patient rating reflecting overall patient satisfaction. The dataset has been uploaded to Hugging Face to facilitate easier access and use by the machine learning community. It contains 161,297 instances in the training set and 53,766 instances in the test set.
### Supported Tasks and Leaderboards
This dataset can be used for sentiment analysis and text classification tasks.
### Languages
The text in the dataset is in English.
## Dataset Structure
### Data Instances
A data instance from the `train` split:
```
{
"drugName": "Buprenex",
"condition": "Pain",
"review": "I have severe drug allergies especially opiates, I have been on Buprenorphine for many years now and found it to be an excellent pain reliever. I have found that at times though it is hard to find and have had to go to the patch.",
"rating": 10,
"date": "May 11, 2012",
"usefulCount": 27
}
```
A data instance from the `test` split:
```
{
"drugName": "Nasacort Allergy 24HR",
"condition": "Allergic Rhinitis",
"review": "Since I start using this product I experienced change of vision and headaches.",
"rating": 3,
"date": "September 8, 2015",
"usefulCount": 27
}
```
#### plain_text
- **Size of the training dataset file:** 82.314 MB
- **Size of the test dataset file:** 27.414 MB
- **Total size of downloaded dataset files:** 109.73 MB
A data instance consists of the following fields:
- `drugName`: The name of the drug reviewed.
- `condition`: The condition for which the drug was prescribed.
- `review`: The text of the review by the patient.
- `rating`: A patient satisfaction rating out of 10.
- `date`: The date when the review was posted.
- `usefulCount`: The number of users who found the review useful.
### Data Fields
- `drugName`: string
- `condition`: string
- `review`: string
- `rating`: integer (0-10)
- `date`: date
- `usefulCount`: integer
### Data Splits
The dataset is split into training and testing sets as follows:
- `train`: 161,297 instances
- `test`: 53,766 instances
## Dataset Creation
### Curation Rationale
The dataset was curated with the intention to study sentiment analysis of drug experience and the transferability of models among different domains and data sources.
### Source Data
#### Initial Data Collection and Normalization
The dataset was collected by crawling online pharmaceutical review sites. No additional preprocessing or normalization has been conducted on the data provided in this repository; it is presented in its original form as obtained from the source.
#### Who are the source language producers?
The reviews were written by patients and users of the drugs.com website.
### Annotations
#### Annotation process
No additional annotation process was followed as the data contains self-reported patient ratings.
#### Who are the annotators?
N/A
### Personal and Sensitive Information
The dataset contains sensitive information in the form of patient drug reviews.
## Considerations for Using the Data
### Social Impact of Dataset
The dataset can be used to study the impact of drugs on patients, which can be beneficial for healthcare research.
### Discussion of Biases
No known biases but users should consider the self-reported nature of the data.
### Other Known Limitations
The dataset may not generalize well to drugs or conditions not represented in the data.
## Additional Information
### Dataset Curators
Curated by the UCI Machine Learning Repository.
### Licensing Information
Licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0).
### Citation Information
When using this dataset, please cite the original source and also the Hugging Face dataset repository:
```
@misc{misc_drug_review_dataset_(drugs.com)_462,
author = {Kallumadi,Surya and Grer,Felix},
title = {Drug Review Dataset (Drugs.com)},
year = {2018},
howpublished = {UCI Machine Learning Repository},
note = {DOI: https://doi.org/10.24432/C5SK5S}
}
```
And:
```
@misc{huggingface:drugscom_reviews,
title = {Drug Review Dataset (Drugs.com) - Hugging Face Version},
author = {Salod, Zakia},
year = {2023},
publisher = {Hugging Face},
howpublished = {Hugging Face Datasets Library},
url = {https://huggingface.co/datasets/Zakia/drugscom_reviews}
}
```
### Contributions
This dataset was uploaded to Hugging Face by [Zakia](https://huggingface.co/Zakia). Special thanks to the community for their interest and engagement. Future contributions, including discussions, issues, and improvements to the dataset card, are welcomed and greatly appreciated in the [discussions section of this dataset's Hugging Face page](https://huggingface.co/datasets/Zakia/drugscom_reviews/discussions).
| Zakia/drugscom_reviews | [
"task_categories:text-classification",
"task_ids:sentiment-classification",
"annotations_creators:no-annotation",
"language_creators:found",
"multilinguality:monolingual",
"size_categories:100K<n<1M",
"source_datasets:original",
"language:en",
"license:cc-by-4.0",
"health",
"medicine",
"patient reviews",
"drug reviews",
"sentiment analysis",
"region:us"
] | 2023-11-30T16:29:12+00:00 | {"annotations_creators": ["no-annotation"], "language_creators": ["found"], "language": ["en"], "license": ["cc-by-4.0"], "multilinguality": ["monolingual"], "size_categories": ["100K<n<1M"], "source_datasets": ["original"], "task_categories": ["text-classification"], "task_ids": ["sentiment-classification"], "pretty_name": "DrugsCom Reviews", "tags": ["health", "medicine", "patient reviews", "drug reviews", "sentiment analysis"]} | 2023-12-11T16:21:20+00:00 | [] | [
"en"
] | TAGS
#task_categories-text-classification #task_ids-sentiment-classification #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-100K<n<1M #source_datasets-original #language-English #license-cc-by-4.0 #health #medicine #patient reviews #drug reviews #sentiment analysis #region-us
|
# Dataset Card for "DrugsCom Reviews"
## Table of Contents
- Dataset Description
- Dataset Summary
- Supported Tasks and Leaderboards
- Languages
- Dataset Structure
- Data Instances
- Data Fields
- Data Splits
- Dataset Creation
- Curation Rationale
- Source Data
- Annotations
- Considerations for Using the Data
- Social Impact of Dataset
- Discussion of Biases
- Other Known Limitations
- Additional Information
- Dataset Curators
- Licensing Information
- Citation Information
- Contributions
## Dataset Description
- Homepage: UCI Machine Learning Repository
- Repository: DrugsCom Reviews on Hugging Face
- Size of the training dataset file: 82.314 MB
- Size of the test dataset file: 27.414 MB
- Total size of downloaded dataset files: 109.73 MB
### Dataset Summary
The DrugsCom Reviews dataset is originally sourced from the UCI Machine Learning Repository. It provides patient reviews on specific drugs along with related conditions and a 10-star patient rating reflecting overall patient satisfaction. The dataset has been uploaded to Hugging Face to facilitate easier access and use by the machine learning community. It contains 161,297 instances in the training set and 53,766 instances in the test set.
### Supported Tasks and Leaderboards
This dataset can be used for sentiment analysis and text classification tasks.
### Languages
The text in the dataset is in English.
## Dataset Structure
### Data Instances
A data instance from the 'train' split:
A data instance from the 'test' split:
#### plain_text
- Size of the training dataset file: 82.314 MB
- Size of the test dataset file: 27.414 MB
- Total size of downloaded dataset files: 109.73 MB
A data instance consists of the following fields:
- 'drugName': The name of the drug reviewed.
- 'condition': The condition for which the drug was prescribed.
- 'review': The text of the review by the patient.
- 'rating': A patient satisfaction rating out of 10.
- 'date': The date when the review was posted.
- 'usefulCount': The number of users who found the review useful.
### Data Fields
- 'drugName': string
- 'condition': string
- 'review': string
- 'rating': integer (0-10)
- 'date': date
- 'usefulCount': integer
### Data Splits
The dataset is split into training and testing sets as follows:
- 'train': 161,297 instances
- 'test': 53,766 instances
## Dataset Creation
### Curation Rationale
The dataset was curated with the intention to study sentiment analysis of drug experience and the transferability of models among different domains and data sources.
### Source Data
#### Initial Data Collection and Normalization
The dataset was collected by crawling online pharmaceutical review sites. No additional preprocessing or normalization has been conducted on the data provided in this repository; it is presented in its original form as obtained from the source.
#### Who are the source language producers?
The reviews were written by patients and users of the URL website.
### Annotations
#### Annotation process
No additional annotation process was followed as the data contains self-reported patient ratings.
#### Who are the annotators?
N/A
### Personal and Sensitive Information
The dataset contains sensitive information in the form of patient drug reviews.
## Considerations for Using the Data
### Social Impact of Dataset
The dataset can be used to study the impact of drugs on patients, which can be beneficial for healthcare research.
### Discussion of Biases
No known biases but users should consider the self-reported nature of the data.
### Other Known Limitations
The dataset may not generalize well to drugs or conditions not represented in the data.
## Additional Information
### Dataset Curators
Curated by the UCI Machine Learning Repository.
### Licensing Information
Licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0).
When using this dataset, please cite the original source and also the Hugging Face dataset repository:
And:
### Contributions
This dataset was uploaded to Hugging Face by Zakia. Special thanks to the community for their interest and engagement. Future contributions, including discussions, issues, and improvements to the dataset card, are welcomed and greatly appreciated in the discussions section of this dataset's Hugging Face page.
| [
"# Dataset Card for \"DrugsCom Reviews\"",
"## 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- 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: UCI Machine Learning Repository\n- Repository: DrugsCom Reviews on Hugging Face\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB",
"### Dataset Summary\n\nThe DrugsCom Reviews dataset is originally sourced from the UCI Machine Learning Repository. It provides patient reviews on specific drugs along with related conditions and a 10-star patient rating reflecting overall patient satisfaction. The dataset has been uploaded to Hugging Face to facilitate easier access and use by the machine learning community. It contains 161,297 instances in the training set and 53,766 instances in the test set.",
"### Supported Tasks and Leaderboards\n\nThis dataset can be used for sentiment analysis and text classification tasks.",
"### Languages\n\nThe text in the dataset is in English.",
"## Dataset Structure",
"### Data Instances\n\nA data instance from the 'train' split:\n\n\n\nA data instance from the 'test' split:",
"#### plain_text\n\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB\n\nA data instance consists of the following fields:\n\n- 'drugName': The name of the drug reviewed.\n- 'condition': The condition for which the drug was prescribed.\n- 'review': The text of the review by the patient.\n- 'rating': A patient satisfaction rating out of 10.\n- 'date': The date when the review was posted.\n- 'usefulCount': The number of users who found the review useful.",
"### Data Fields\n\n- 'drugName': string\n- 'condition': string\n- 'review': string\n- 'rating': integer (0-10)\n- 'date': date\n- 'usefulCount': integer",
"### Data Splits\n\nThe dataset is split into training and testing sets as follows:\n\n- 'train': 161,297 instances\n- 'test': 53,766 instances",
"## Dataset Creation",
"### Curation Rationale\n\nThe dataset was curated with the intention to study sentiment analysis of drug experience and the transferability of models among different domains and data sources.",
"### Source Data",
"#### Initial Data Collection and Normalization\n\nThe dataset was collected by crawling online pharmaceutical review sites. No additional preprocessing or normalization has been conducted on the data provided in this repository; it is presented in its original form as obtained from the source.",
"#### Who are the source language producers?\n\nThe reviews were written by patients and users of the URL website.",
"### Annotations",
"#### Annotation process\n\nNo additional annotation process was followed as the data contains self-reported patient ratings.",
"#### Who are the annotators?\n\nN/A",
"### Personal and Sensitive Information\n\nThe dataset contains sensitive information in the form of patient drug reviews.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe dataset can be used to study the impact of drugs on patients, which can be beneficial for healthcare research.",
"### Discussion of Biases\n\nNo known biases but users should consider the self-reported nature of the data.",
"### Other Known Limitations\n\nThe dataset may not generalize well to drugs or conditions not represented in the data.",
"## Additional Information",
"### Dataset Curators\n\nCurated by the UCI Machine Learning Repository.",
"### Licensing Information\n\nLicensed under the Creative Commons Attribution 4.0 International (CC BY 4.0).\n\n\n\n\nWhen using this dataset, please cite the original source and also the Hugging Face dataset repository:\n\n\n\nAnd:",
"### Contributions\n\nThis dataset was uploaded to Hugging Face by Zakia. Special thanks to the community for their interest and engagement. Future contributions, including discussions, issues, and improvements to the dataset card, are welcomed and greatly appreciated in the discussions section of this dataset's Hugging Face page."
] | [
"TAGS\n#task_categories-text-classification #task_ids-sentiment-classification #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-100K<n<1M #source_datasets-original #language-English #license-cc-by-4.0 #health #medicine #patient reviews #drug reviews #sentiment analysis #region-us \n",
"# Dataset Card for \"DrugsCom Reviews\"",
"## 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- 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: UCI Machine Learning Repository\n- Repository: DrugsCom Reviews on Hugging Face\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB",
"### Dataset Summary\n\nThe DrugsCom Reviews dataset is originally sourced from the UCI Machine Learning Repository. It provides patient reviews on specific drugs along with related conditions and a 10-star patient rating reflecting overall patient satisfaction. The dataset has been uploaded to Hugging Face to facilitate easier access and use by the machine learning community. It contains 161,297 instances in the training set and 53,766 instances in the test set.",
"### Supported Tasks and Leaderboards\n\nThis dataset can be used for sentiment analysis and text classification tasks.",
"### Languages\n\nThe text in the dataset is in English.",
"## Dataset Structure",
"### Data Instances\n\nA data instance from the 'train' split:\n\n\n\nA data instance from the 'test' split:",
"#### plain_text\n\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB\n\nA data instance consists of the following fields:\n\n- 'drugName': The name of the drug reviewed.\n- 'condition': The condition for which the drug was prescribed.\n- 'review': The text of the review by the patient.\n- 'rating': A patient satisfaction rating out of 10.\n- 'date': The date when the review was posted.\n- 'usefulCount': The number of users who found the review useful.",
"### Data Fields\n\n- 'drugName': string\n- 'condition': string\n- 'review': string\n- 'rating': integer (0-10)\n- 'date': date\n- 'usefulCount': integer",
"### Data Splits\n\nThe dataset is split into training and testing sets as follows:\n\n- 'train': 161,297 instances\n- 'test': 53,766 instances",
"## Dataset Creation",
"### Curation Rationale\n\nThe dataset was curated with the intention to study sentiment analysis of drug experience and the transferability of models among different domains and data sources.",
"### Source Data",
"#### Initial Data Collection and Normalization\n\nThe dataset was collected by crawling online pharmaceutical review sites. No additional preprocessing or normalization has been conducted on the data provided in this repository; it is presented in its original form as obtained from the source.",
"#### Who are the source language producers?\n\nThe reviews were written by patients and users of the URL website.",
"### Annotations",
"#### Annotation process\n\nNo additional annotation process was followed as the data contains self-reported patient ratings.",
"#### Who are the annotators?\n\nN/A",
"### Personal and Sensitive Information\n\nThe dataset contains sensitive information in the form of patient drug reviews.",
"## Considerations for Using the Data",
"### Social Impact of Dataset\n\nThe dataset can be used to study the impact of drugs on patients, which can be beneficial for healthcare research.",
"### Discussion of Biases\n\nNo known biases but users should consider the self-reported nature of the data.",
"### Other Known Limitations\n\nThe dataset may not generalize well to drugs or conditions not represented in the data.",
"## Additional Information",
"### Dataset Curators\n\nCurated by the UCI Machine Learning Repository.",
"### Licensing Information\n\nLicensed under the Creative Commons Attribution 4.0 International (CC BY 4.0).\n\n\n\n\nWhen using this dataset, please cite the original source and also the Hugging Face dataset repository:\n\n\n\nAnd:",
"### Contributions\n\nThis dataset was uploaded to Hugging Face by Zakia. Special thanks to the community for their interest and engagement. Future contributions, including discussions, issues, and improvements to the dataset card, are welcomed and greatly appreciated in the discussions section of this dataset's Hugging Face page."
] | [
105,
13,
113,
68,
99,
26,
14,
6,
27,
142,
49,
41,
5,
37,
4,
65,
23,
5,
25,
12,
23,
8,
32,
27,
26,
5,
18,
45,
73
] | [
"passage: TAGS\n#task_categories-text-classification #task_ids-sentiment-classification #annotations_creators-no-annotation #language_creators-found #multilinguality-monolingual #size_categories-100K<n<1M #source_datasets-original #language-English #license-cc-by-4.0 #health #medicine #patient reviews #drug reviews #sentiment analysis #region-us \n# Dataset Card for \"DrugsCom Reviews\"## 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- 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: UCI Machine Learning Repository\n- Repository: DrugsCom Reviews on Hugging Face\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB### Dataset Summary\n\nThe DrugsCom Reviews dataset is originally sourced from the UCI Machine Learning Repository. It provides patient reviews on specific drugs along with related conditions and a 10-star patient rating reflecting overall patient satisfaction. The dataset has been uploaded to Hugging Face to facilitate easier access and use by the machine learning community. It contains 161,297 instances in the training set and 53,766 instances in the test set.### Supported Tasks and Leaderboards\n\nThis dataset can be used for sentiment analysis and text classification tasks.### Languages\n\nThe text in the dataset is in English.## Dataset Structure### Data Instances\n\nA data instance from the 'train' split:\n\n\n\nA data instance from the 'test' split:",
"passage: #### plain_text\n\n- Size of the training dataset file: 82.314 MB\n- Size of the test dataset file: 27.414 MB\n- Total size of downloaded dataset files: 109.73 MB\n\nA data instance consists of the following fields:\n\n- 'drugName': The name of the drug reviewed.\n- 'condition': The condition for which the drug was prescribed.\n- 'review': The text of the review by the patient.\n- 'rating': A patient satisfaction rating out of 10.\n- 'date': The date when the review was posted.\n- 'usefulCount': The number of users who found the review useful.### Data Fields\n\n- 'drugName': string\n- 'condition': string\n- 'review': string\n- 'rating': integer (0-10)\n- 'date': date\n- 'usefulCount': integer### Data Splits\n\nThe dataset is split into training and testing sets as follows:\n\n- 'train': 161,297 instances\n- 'test': 53,766 instances## Dataset Creation### Curation Rationale\n\nThe dataset was curated with the intention to study sentiment analysis of drug experience and the transferability of models among different domains and data sources.### Source Data#### Initial Data Collection and Normalization\n\nThe dataset was collected by crawling online pharmaceutical review sites. No additional preprocessing or normalization has been conducted on the data provided in this repository; it is presented in its original form as obtained from the source.#### Who are the source language producers?\n\nThe reviews were written by patients and users of the URL website.### Annotations#### Annotation process\n\nNo additional annotation process was followed as the data contains self-reported patient ratings.#### Who are the annotators?\n\nN/A### Personal and Sensitive Information\n\nThe dataset contains sensitive information in the form of patient drug reviews.## Considerations for Using the Data### Social Impact of Dataset\n\nThe dataset can be used to study the impact of drugs on patients, which can be beneficial for healthcare research.### Discussion of Biases\n\nNo known biases but users should consider the self-reported nature of the data.### Other Known Limitations\n\nThe dataset may not generalize well to drugs or conditions not represented in the data.## Additional Information### Dataset Curators\n\nCurated by the UCI Machine Learning Repository.### Licensing Information\n\nLicensed under the Creative Commons Attribution 4.0 International (CC BY 4.0).\n\n\n\n\nWhen using this dataset, please cite the original source and also the Hugging Face dataset repository:\n\n\n\nAnd:"
] |
bccadcf308dc5496690d2ed1bd0572cd5c3fbccc | # Dataset Card for "autotrain-data-reformatted_test"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | healthcorum/autotrain-data-reformatted_test | [
"region:us"
] | 2023-11-30T17:07:12+00:00 | {"dataset_info": {"features": [{"name": "Unnamed: 0", "dtype": "int64"}, {"name": "autotrain_text", "dtype": "string"}, {"name": "autotrain_label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 9500675, "num_examples": 7998}, {"name": "validation", "num_bytes": 2375962, "num_examples": 2000}], "download_size": 4169235, "dataset_size": 11876637}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}]}]} | 2023-11-30T17:07:18+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "autotrain-data-reformatted_test"
More Information needed | [
"# Dataset Card for \"autotrain-data-reformatted_test\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"autotrain-data-reformatted_test\"\n\nMore Information needed"
] | [
6,
21
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"autotrain-data-reformatted_test\"\n\nMore Information needed"
] |
c2f2198de175b11cad71d4fd934832c4ee531377 | # No Prompt
This is a dataset created to test language models on generating high-quality, useful text without prompt formatting. This works by simply removing the formatting from the dataset to be used, be it guanaco, openassistant, etc... | appvoid/no-prompt-15k | [
"region:us"
] | 2023-11-30T17:16:57+00:00 | {"dataset_info": {"features": [{"name": "instruction", "dtype": "string"}, {"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}, {"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 37820576, "num_examples": 15000}], "download_size": 20067913, "dataset_size": 37820576}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-05T13:05:08+00:00 | [] | [] | TAGS
#region-us
| # No Prompt
This is a dataset created to test language models on generating high-quality, useful text without prompt formatting. This works by simply removing the formatting from the dataset to be used, be it guanaco, openassistant, etc... | [
"# No Prompt\n\nThis is a dataset created to test language models on generating high-quality, useful text without prompt formatting. This works by simply removing the formatting from the dataset to be used, be it guanaco, openassistant, etc..."
] | [
"TAGS\n#region-us \n",
"# No Prompt\n\nThis is a dataset created to test language models on generating high-quality, useful text without prompt formatting. This works by simply removing the formatting from the dataset to be used, be it guanaco, openassistant, etc..."
] | [
6,
58
] | [
"passage: TAGS\n#region-us \n# No Prompt\n\nThis is a dataset created to test language models on generating high-quality, useful text without prompt formatting. This works by simply removing the formatting from the dataset to be used, be it guanaco, openassistant, etc..."
] |
6883c8820fd21a4af731118b2e0b3d293e439d77 |
# Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed] | siqideng/proposal_drafter_feedback | [
"region:us"
] | 2023-11-30T17:19:30+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data.csv"}]}]} | 2023-12-20T15:07:07+00:00 | [] | [] | TAGS
#region-us
|
# Dataset Card for Dataset Name
## Dataset Details
### Dataset Description
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
| [
"# Dataset Card for Dataset Name",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for Dataset Name",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
6,
8,
4,
40,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] | [
"passage: TAGS\n#region-us \n# Dataset Card for Dataset Name## Dataset Details### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
1366e3fc86adf8b30b7cac109905dc3fcb0026e0 | # Dataset Card for "elephants_512"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | mespinosami/elephants_512 | [
"region:us"
] | 2023-11-30T17:23:23+00:00 | {"dataset_info": {"features": [{"name": "image", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 845894902.629, "num_examples": 9023}], "download_size": 840684586, "dataset_size": 845894902.629}} | 2023-11-30T17:27:55+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "elephants_512"
More Information needed | [
"# Dataset Card for \"elephants_512\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"elephants_512\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"elephants_512\"\n\nMore Information needed"
] |
6ee48a7406956528f495084bcc643bbe7eae49ab |
# Bangumi Image Base of Akatsuki No Yona
This is the image base of bangumi Akatsuki no Yona, we detected 41 characters, 3412 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 532 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 33 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 76 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 69 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 39 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 34 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 18 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 213 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 46 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 207 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 29 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 58 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 50 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 60 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 35 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 58 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 28 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 15 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 15 | [Download](18/dataset.zip) |  |  |  |  |  |  |  |  |
| 19 | 230 | [Download](19/dataset.zip) |  |  |  |  |  |  |  |  |
| 20 | 57 | [Download](20/dataset.zip) |  |  |  |  |  |  |  |  |
| 21 | 22 | [Download](21/dataset.zip) |  |  |  |  |  |  |  |  |
| 22 | 85 | [Download](22/dataset.zip) |  |  |  |  |  |  |  |  |
| 23 | 31 | [Download](23/dataset.zip) |  |  |  |  |  |  |  |  |
| 24 | 21 | [Download](24/dataset.zip) |  |  |  |  |  |  |  |  |
| 25 | 25 | [Download](25/dataset.zip) |  |  |  |  |  |  |  |  |
| 26 | 9 | [Download](26/dataset.zip) |  |  |  |  |  |  |  |  |
| 27 | 21 | [Download](27/dataset.zip) |  |  |  |  |  |  |  |  |
| 28 | 797 | [Download](28/dataset.zip) |  |  |  |  |  |  |  |  |
| 29 | 77 | [Download](29/dataset.zip) |  |  |  |  |  |  |  |  |
| 30 | 11 | [Download](30/dataset.zip) |  |  |  |  |  |  |  |  |
| 31 | 7 | [Download](31/dataset.zip) |  |  |  |  |  |  |  | N/A |
| 32 | 14 | [Download](32/dataset.zip) |  |  |  |  |  |  |  |  |
| 33 | 26 | [Download](33/dataset.zip) |  |  |  |  |  |  |  |  |
| 34 | 41 | [Download](34/dataset.zip) |  |  |  |  |  |  |  |  |
| 35 | 14 | [Download](35/dataset.zip) |  |  |  |  |  |  |  |  |
| 36 | 6 | [Download](36/dataset.zip) |  |  |  |  |  |  | N/A | N/A |
| 37 | 14 | [Download](37/dataset.zip) |  |  |  |  |  |  |  |  |
| 38 | 9 | [Download](38/dataset.zip) |  |  |  |  |  |  |  |  |
| 39 | 46 | [Download](39/dataset.zip) |  |  |  |  |  |  |  |  |
| noise | 234 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/akatsukinoyona | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T17:36:17+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T19:21:29+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of Akatsuki No Yona
======================================
This is the image base of bangumi Akatsuki no Yona, we detected 41 characters, 3412 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
2cff1bd59cfb92093cb403286fe6f344eb097bf6 |
# Bangumi Image Base of Mairimashita! Iruma-kun
This is the image base of bangumi Mairimashita! Iruma-kun, we detected 94 characters, 8676 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 1627 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 414 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 36 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 70 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 100 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 19 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 73 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 159 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 31 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 27 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 21 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 255 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 35 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 297 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 50 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 34 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 15 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 36 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 28 | [Download](18/dataset.zip) |  |  |  |  |  |  |  |  |
| 19 | 156 | [Download](19/dataset.zip) |  |  |  |  |  |  |  |  |
| 20 | 22 | [Download](20/dataset.zip) |  |  |  |  |  |  |  |  |
| 21 | 31 | [Download](21/dataset.zip) |  |  |  |  |  |  |  |  |
| 22 | 16 | [Download](22/dataset.zip) |  |  |  |  |  |  |  |  |
| 23 | 65 | [Download](23/dataset.zip) |  |  |  |  |  |  |  |  |
| 24 | 70 | [Download](24/dataset.zip) |  |  |  |  |  |  |  |  |
| 25 | 38 | [Download](25/dataset.zip) |  |  |  |  |  |  |  |  |
| 26 | 52 | [Download](26/dataset.zip) |  |  |  |  |  |  |  |  |
| 27 | 731 | [Download](27/dataset.zip) |  |  |  |  |  |  |  |  |
| 28 | 100 | [Download](28/dataset.zip) |  |  |  |  |  |  |  |  |
| 29 | 57 | [Download](29/dataset.zip) |  |  |  |  |  |  |  |  |
| 30 | 88 | [Download](30/dataset.zip) |  |  |  |  |  |  |  |  |
| 31 | 40 | [Download](31/dataset.zip) |  |  |  |  |  |  |  |  |
| 32 | 8 | [Download](32/dataset.zip) |  |  |  |  |  |  |  |  |
| 33 | 23 | [Download](33/dataset.zip) |  |  |  |  |  |  |  |  |
| 34 | 17 | [Download](34/dataset.zip) |  |  |  |  |  |  |  |  |
| 35 | 249 | [Download](35/dataset.zip) |  |  |  |  |  |  |  |  |
| 36 | 171 | [Download](36/dataset.zip) |  |  |  |  |  |  |  |  |
| 37 | 90 | [Download](37/dataset.zip) |  |  |  |  |  |  |  |  |
| 38 | 26 | [Download](38/dataset.zip) |  |  |  |  |  |  |  |  |
| 39 | 36 | [Download](39/dataset.zip) |  |  |  |  |  |  |  |  |
| 40 | 12 | [Download](40/dataset.zip) |  |  |  |  |  |  |  |  |
| 41 | 123 | [Download](41/dataset.zip) |  |  |  |  |  |  |  |  |
| 42 | 52 | [Download](42/dataset.zip) |  |  |  |  |  |  |  |  |
| 43 | 51 | [Download](43/dataset.zip) |  |  |  |  |  |  |  |  |
| 44 | 18 | [Download](44/dataset.zip) |  |  |  |  |  |  |  |  |
| 45 | 15 | [Download](45/dataset.zip) |  |  |  |  |  |  |  |  |
| 46 | 73 | [Download](46/dataset.zip) |  |  |  |  |  |  |  |  |
| 47 | 35 | [Download](47/dataset.zip) |  |  |  |  |  |  |  |  |
| 48 | 22 | [Download](48/dataset.zip) |  |  |  |  |  |  |  |  |
| 49 | 20 | [Download](49/dataset.zip) |  |  |  |  |  |  |  |  |
| 50 | 21 | [Download](50/dataset.zip) |  |  |  |  |  |  |  |  |
| 51 | 8 | [Download](51/dataset.zip) |  |  |  |  |  |  |  |  |
| 52 | 35 | [Download](52/dataset.zip) |  |  |  |  |  |  |  |  |
| 53 | 49 | [Download](53/dataset.zip) |  |  |  |  |  |  |  |  |
| 54 | 31 | [Download](54/dataset.zip) |  |  |  |  |  |  |  |  |
| 55 | 715 | [Download](55/dataset.zip) |  |  |  |  |  |  |  |  |
| 56 | 20 | [Download](56/dataset.zip) |  |  |  |  |  |  |  |  |
| 57 | 153 | [Download](57/dataset.zip) |  |  |  |  |  |  |  |  |
| 58 | 34 | [Download](58/dataset.zip) |  |  |  |  |  |  |  |  |
| 59 | 11 | [Download](59/dataset.zip) |  |  |  |  |  |  |  |  |
| 60 | 34 | [Download](60/dataset.zip) |  |  |  |  |  |  |  |  |
| 61 | 64 | [Download](61/dataset.zip) |  |  |  |  |  |  |  |  |
| 62 | 15 | [Download](62/dataset.zip) |  |  |  |  |  |  |  |  |
| 63 | 63 | [Download](63/dataset.zip) |  |  |  |  |  |  |  |  |
| 64 | 42 | [Download](64/dataset.zip) |  |  |  |  |  |  |  |  |
| 65 | 40 | [Download](65/dataset.zip) |  |  |  |  |  |  |  |  |
| 66 | 17 | [Download](66/dataset.zip) |  |  |  |  |  |  |  |  |
| 67 | 10 | [Download](67/dataset.zip) |  |  |  |  |  |  |  |  |
| 68 | 13 | [Download](68/dataset.zip) |  |  |  |  |  |  |  |  |
| 69 | 22 | [Download](69/dataset.zip) |  |  |  |  |  |  |  |  |
| 70 | 20 | [Download](70/dataset.zip) |  |  |  |  |  |  |  |  |
| 71 | 8 | [Download](71/dataset.zip) |  |  |  |  |  |  |  |  |
| 72 | 9 | [Download](72/dataset.zip) |  |  |  |  |  |  |  |  |
| 73 | 164 | [Download](73/dataset.zip) |  |  |  |  |  |  |  |  |
| 74 | 73 | [Download](74/dataset.zip) |  |  |  |  |  |  |  |  |
| 75 | 40 | [Download](75/dataset.zip) |  |  |  |  |  |  |  |  |
| 76 | 324 | [Download](76/dataset.zip) |  |  |  |  |  |  |  |  |
| 77 | 36 | [Download](77/dataset.zip) |  |  |  |  |  |  |  |  |
| 78 | 11 | [Download](78/dataset.zip) |  |  |  |  |  |  |  |  |
| 79 | 12 | [Download](79/dataset.zip) |  |  |  |  |  |  |  |  |
| 80 | 16 | [Download](80/dataset.zip) |  |  |  |  |  |  |  |  |
| 81 | 8 | [Download](81/dataset.zip) |  |  |  |  |  |  |  |  |
| 82 | 13 | [Download](82/dataset.zip) |  |  |  |  |  |  |  |  |
| 83 | 13 | [Download](83/dataset.zip) |  |  |  |  |  |  |  |  |
| 84 | 14 | [Download](84/dataset.zip) |  |  |  |  |  |  |  |  |
| 85 | 6 | [Download](85/dataset.zip) |  |  |  |  |  |  | N/A | N/A |
| 86 | 16 | [Download](86/dataset.zip) |  |  |  |  |  |  |  |  |
| 87 | 41 | [Download](87/dataset.zip) |  |  |  |  |  |  |  |  |
| 88 | 16 | [Download](88/dataset.zip) |  |  |  |  |  |  |  |  |
| 89 | 13 | [Download](89/dataset.zip) |  |  |  |  |  |  |  |  |
| 90 | 30 | [Download](90/dataset.zip) |  |  |  |  |  |  |  |  |
| 91 | 7 | [Download](91/dataset.zip) |  |  |  |  |  |  |  | N/A |
| 92 | 6 | [Download](92/dataset.zip) |  |  |  |  |  |  | N/A | N/A |
| noise | 549 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/mairimashitairumakun | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T17:37:20+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T21:44:09+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of Mairimashita! Iruma-kun
=============================================
This is the image base of bangumi Mairimashita! Iruma-kun, we detected 94 characters, 8676 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
c2431e4ded6507a753a94574f8157915ec791dbe | # [doc] image dataset 2
This dataset contains 4 jpeg files in the images/ subdirectory. | datasets-examples/doc-image-2 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T17:41:37+00:00 | {"size_categories": ["n<1K"]} | 2023-11-30T17:43:22+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 2
This dataset contains 4 jpeg files in the images/ subdirectory. | [
"# [doc] image dataset 2\n\nThis dataset contains 4 jpeg files in the images/ subdirectory."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 2\n\nThis dataset contains 4 jpeg files in the images/ subdirectory."
] | [
16,
25
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 2\n\nThis dataset contains 4 jpeg files in the images/ subdirectory."
] |
efdc57f60bd53acf991b7ae01138ca1a4221e25f | # [doc] image dataset 3
This dataset contains 4 image files with different formats in the images/ subdirectory. | datasets-examples/doc-image-3 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T17:45:50+00:00 | {"size_categories": ["n<1K"]} | 2023-11-30T17:47:28+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 3
This dataset contains 4 image files with different formats in the images/ subdirectory. | [
"# [doc] image dataset 3\n\nThis dataset contains 4 image files with different formats in the images/ subdirectory."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 3\n\nThis dataset contains 4 image files with different formats in the images/ subdirectory."
] | [
16,
28
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 3\n\nThis dataset contains 4 image files with different formats in the images/ subdirectory."
] |
04a61ed6f5ef2680e309838d6a6fc5e4f3c8cc79 | # [doc] image dataset 4
This dataset contains 4 jpeg files in the `train/` subdirectory, along with a `metadata.csv` file that provides the data for other columns. | datasets-examples/doc-image-4 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T17:50:03+00:00 | {"size_categories": ["n<1K"]} | 2024-01-05T18:18:38+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 4
This dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns. | [
"# [doc] image dataset 4\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 4\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] | [
16,
45
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 4\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] |
f955e04f6d13c367a711a0a07571a856cb1077fb | # Dataset Card for "hubert_layer9-librispeech-asr100h_tokenized_final_asr"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | cmu-mlsp/hubert_layer9-librispeech-asr100h_tokenized_final_asr | [
"region:us"
] | 2023-11-30T17:50:40+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}, {"split": "validation_tts", "path": "data/validation_tts-*"}, {"split": "test", "path": "data/test-*"}, {"split": "test_tts", "path": "data/test_tts-*"}]}], "dataset_info": {"features": [{"name": "input_ids", "sequence": "int32"}, {"name": "attention_mask", "sequence": "int8"}, {"name": "labels", "sequence": "int64"}], "splits": [{"name": "train", "num_bytes": 503610155, "num_examples": 28539}, {"name": "validation", "num_bytes": 26404561, "num_examples": 2703}, {"name": "validation_tts", "num_bytes": 26404561, "num_examples": 2703}, {"name": "test", "num_bytes": 26173400, "num_examples": 2620}, {"name": "test_tts", "num_bytes": 26173400, "num_examples": 2620}], "download_size": 62611869, "dataset_size": 608766077}} | 2023-11-30T17:50:49+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "hubert_layer9-librispeech-asr100h_tokenized_final_asr"
More Information needed | [
"# Dataset Card for \"hubert_layer9-librispeech-asr100h_tokenized_final_asr\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"hubert_layer9-librispeech-asr100h_tokenized_final_asr\"\n\nMore Information needed"
] | [
6,
34
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"hubert_layer9-librispeech-asr100h_tokenized_final_asr\"\n\nMore Information needed"
] |
4d9ad2e222331d4a12d12fa6ef6387015c6a8c73 | # [doc] image dataset 5
This dataset contains 4 jpeg files in the `train/` subdirectory, along with a `metadata.jsonl` file that provides the data for other columns. | datasets-examples/doc-image-5 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T17:53:58+00:00 | {"size_categories": ["n<1K"]} | 2024-01-05T18:18:55+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 5
This dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns. | [
"# [doc] image dataset 5\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 5\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] | [
16,
45
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 5\n\nThis dataset contains 4 jpeg files in the 'train/' subdirectory, along with a 'URL' file that provides the data for other columns."
] |
a62590f822434131aaa7aaeaee2f39304427b842 | # Dataset Card for "biosift"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | AshtonIsNotHere/biosift | [
"region:us"
] | 2023-11-30T17:59:53+00:00 | {"dataset_info": {"features": [{"name": "PMID", "dtype": "int64"}, {"name": "Title", "dtype": "string"}, {"name": "Abstract", "dtype": "string"}, {"name": "Split", "dtype": "string"}, {"name": "Number of Annotators", "dtype": "int64"}, {"name": "Aggregate", "dtype": "int64"}, {"name": "Has Human Subjects", "dtype": "float64"}, {"name": "Has Target Disease", "dtype": "float64"}, {"name": "Cohort Study or Clinical Trial", "dtype": "float64"}, {"name": "Has Quantitative Outcome Measure", "dtype": "float64"}, {"name": "Has Study Drug(s)", "dtype": "float64"}, {"name": "Has Population Size", "dtype": "float64"}, {"name": "Has Comparator Group", "dtype": "float64"}, {"name": "label", "sequence": "float64"}, {"name": "__index_level_0__", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 15286088, "num_examples": 8005}, {"name": "validation", "num_bytes": 1931610, "num_examples": 997}, {"name": "test", "num_bytes": 1923714, "num_examples": 998}], "download_size": 9802250, "dataset_size": 19141412}} | 2023-11-30T18:00:28+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "biosift"
More Information needed | [
"# Dataset Card for \"biosift\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"biosift\"\n\nMore Information needed"
] | [
6,
13
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"biosift\"\n\nMore Information needed"
] |
5a6e5b8f6e0bdbb3353962903d3115c2463a78f9 | # Dataset Card for "openai_summarize_comparisons_tldrprompt_relabel1b"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | mnoukhov/openai_summarize_comparisons_tldrprompt_relabel1b | [
"region:us"
] | 2023-11-30T18:14:46+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "test", "path": "data/test-*"}]}], "dataset_info": {"features": [{"name": "prompt", "dtype": "string"}, {"name": "chosen", "dtype": "string"}, {"name": "rejected", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 156593160, "num_examples": 92534}, {"name": "test", "num_bytes": 8322345, "num_examples": 5000}], "download_size": 21793816, "dataset_size": 164915505}} | 2023-12-02T22:00:36+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "openai_summarize_comparisons_tldrprompt_relabel1b"
More Information needed | [
"# Dataset Card for \"openai_summarize_comparisons_tldrprompt_relabel1b\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"openai_summarize_comparisons_tldrprompt_relabel1b\"\n\nMore Information needed"
] | [
6,
32
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"openai_summarize_comparisons_tldrprompt_relabel1b\"\n\nMore Information needed"
] |
ec03afeacb2c01bcc85c09ed33d9e03b386a2685 | # Dataset Card for "ultrafeedback_binarized"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | rajammanabrolu/ultrafeedback_binarized | [
"region:us"
] | 2023-11-30T18:23:24+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train_prefs", "path": "data/train_prefs-*"}, {"split": "train_sft", "path": "data/train_sft-*"}, {"split": "test_prefs", "path": "data/test_prefs-*"}, {"split": "test_sft", "path": "data/test_sft-*"}, {"split": "train_gen", "path": "data/train_gen-*"}, {"split": "test_gen", "path": "data/test_gen-*"}]}], "dataset_info": {"features": [{"name": "prompt", "dtype": "string"}, {"name": "prompt_id", "dtype": "string"}, {"name": "chosen", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "rejected", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "messages", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "score_chosen", "dtype": "float64"}, {"name": "score_rejected", "dtype": "float64"}], "splits": [{"name": "train_prefs", "num_bytes": 397273717, "num_examples": 61966}, {"name": "train_sft", "num_bytes": 397273717, "num_examples": 61966}, {"name": "test_prefs", "num_bytes": 12782225, "num_examples": 2000}, {"name": "test_sft", "num_bytes": 6270496, "num_examples": 1000}, {"name": "train_gen", "num_bytes": 316634390, "num_examples": 61966}, {"name": "test_gen", "num_bytes": 5008220, "num_examples": 1000}], "download_size": 636473621, "dataset_size": 1135242765}} | 2023-11-30T18:23:52+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "ultrafeedback_binarized"
More Information needed | [
"# Dataset Card for \"ultrafeedback_binarized\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"ultrafeedback_binarized\"\n\nMore Information needed"
] | [
6,
18
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"ultrafeedback_binarized\"\n\nMore Information needed"
] |
d86b0e366f0f4d4ec4df273e43d28015289ca0d8 | # Dataset Card for "ukioye_editing"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | tianyi0216/ukioye_editing | [
"region:us"
] | 2023-11-30T18:23:46+00:00 | {"dataset_info": {"features": [{"name": "source_img", "dtype": "image"}, {"name": "instruction", "dtype": "string"}, {"name": "target_img", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 447007253.829, "num_examples": 1989}], "download_size": 446312410, "dataset_size": 447007253.829}} | 2023-12-01T01:10:40+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "ukioye_editing"
More Information needed | [
"# Dataset Card for \"ukioye_editing\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"ukioye_editing\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"ukioye_editing\"\n\nMore Information needed"
] |
d3621480d0523d10966db3d64671dca546a3c7a5 |
# Dataset Card for Hugging Face Hub Model Cards with Embeddings
This dataset consists of [model cards](https://huggingface.co/docs/hub/model-cards) for models hosted on the Hugging Face Hub. The model cards are created by the community and provide information about the model, its performance, its intended uses, and more.
This dataset is updated on a daily basis and includes publicly available models on the Hugging Face Hub.
This dataset is made available to help support users wanting to work with a large number of Model Cards from the Hub. We hope that this dataset will help support research in the area of Model Cards and their use but the format of this dataset may not be useful for all use cases. If there are other features that you would like to see included in this dataset, please open a new [discussion](https://huggingface.co/datasets/librarian-bots/model_cards_with_metadata/discussions/new).
This dataset is the same as the [Hugging Face Hub Model Cards](https://huggingface.co/datasets/librarian-bots/model_cards) dataset but with the addition of embeddings for each model card. The embeddings are generated using the [jinaai/jina-embeddings-v2-base-en](https://huggingface.co/jinaai/jina-embeddings-v2-base-en) model.
## Dataset Details
### Dataset Description
- **Curated by:** Daniel van Strien
- **Language(s) (NLP):** Model cards on the Hugging Face Hub are predominantly in English but may include other languages.
## Uses
There are a number of potential uses for this dataset including:
- text mining to find common themes in model cards
- analysis of the model card format/content
- topic modelling of model cards
- analysis of the model card metadata
- training language models on model cards
- build a recommender system for model cards
- build a search engine for model cards
### Out-of-Scope Use
[More Information Needed]
## Dataset Structure
This dataset has a single split.
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
The dataset was created to assist people in working with model cards. In particular. it was created to support research in the area of model cards and their use. It is possible to use the Hugging Face Hub API or client library to download model cards and this option may be preferable if you have a very specific use case or require a different format.
### Source Data
The source data is `README.md` files for models hosted on the Hugging Face Hub. We do not include any other supplementary files that may be included in the model card directory.
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
The data is downloaded using a CRON job on a daily basis.
#### Who are the source data producers?
The source data producers are the creators of the model cards on the Hugging Face Hub. This includes a broad variety of people from the community ranging from large companies to individual researchers. We do not gather any information about who created the model card in this repository although this information can be gathered from the Hugging Face Hub API.
### Annotations [optional]
There are no additional annotations in this dataset beyond the model card content.
#### Annotation process
N/A
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
N/A
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
We make no effort to anonymize the data. Whilst we don't expect the majority of model cards to contain personal or sensitive information, it is possible that some model cards may contain this information. Model cards may also link to websites or email addresses.
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
Model cards are created by the community and we do not have any control over the content of the model cards. We do not review the content of the model cards and we do not make any claims about the accuracy of the information in the model cards.
Some model cards will themselves discuss bias and sometimes this is done by providing examples of bias in either the training data or the responses provided by the model. As a result this dataset may contain examples of bias.
Whilst we do not directly download any images linked to the model cards, some model cards may include images. Some of these images may not be suitable for all audiences.
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
## Citation
No formal citation is required for this dataset but if you use this dataset in your work, please include a link to this dataset page.
## Dataset Card Authors
[@davanstrien](https://huggingface.co/davanstrien)
## Dataset Card Contact
[@davanstrien](https://huggingface.co/davanstrien) | librarian-bots/model_cards_with_metadata_with_embeddings | [
"task_categories:text-retrieval",
"size_categories:100K<n<1M",
"region:us"
] | 2023-11-30T18:37:44+00:00 | {"size_categories": ["100K<n<1M"], "task_categories": ["text-retrieval"], "pretty_name": "Model Card", "dataset_info": {"features": [{"name": "modelId", "dtype": "string"}, {"name": "author", "dtype": "string"}, {"name": "last_modified", "dtype": "timestamp[us, tz=UTC]"}, {"name": "downloads", "dtype": "int64"}, {"name": "likes", "dtype": "int64"}, {"name": "library_name", "dtype": "string"}, {"name": "tags", "sequence": "string"}, {"name": "pipeline_tag", "dtype": "string"}, {"name": "createdAt", "dtype": "timestamp[us, tz=UTC]"}, {"name": "card", "dtype": "string"}, {"name": "embedding", "sequence": "float32"}], "splits": [{"name": "train", "num_bytes": 2104883666.6585019, "num_examples": 442651}], "download_size": 1243809305, "dataset_size": 2104883666.6585019}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-18T10:15:58+00:00 | [] | [] | TAGS
#task_categories-text-retrieval #size_categories-100K<n<1M #region-us
|
# Dataset Card for Hugging Face Hub Model Cards with Embeddings
This dataset consists of model cards for models hosted on the Hugging Face Hub. The model cards are created by the community and provide information about the model, its performance, its intended uses, and more.
This dataset is updated on a daily basis and includes publicly available models on the Hugging Face Hub.
This dataset is made available to help support users wanting to work with a large number of Model Cards from the Hub. We hope that this dataset will help support research in the area of Model Cards and their use but the format of this dataset may not be useful for all use cases. If there are other features that you would like to see included in this dataset, please open a new discussion.
This dataset is the same as the Hugging Face Hub Model Cards dataset but with the addition of embeddings for each model card. The embeddings are generated using the jinaai/jina-embeddings-v2-base-en model.
## Dataset Details
### Dataset Description
- Curated by: Daniel van Strien
- Language(s) (NLP): Model cards on the Hugging Face Hub are predominantly in English but may include other languages.
## Uses
There are a number of potential uses for this dataset including:
- text mining to find common themes in model cards
- analysis of the model card format/content
- topic modelling of model cards
- analysis of the model card metadata
- training language models on model cards
- build a recommender system for model cards
- build a search engine for model cards
### Out-of-Scope Use
## Dataset Structure
This dataset has a single split.
## Dataset Creation
### Curation Rationale
The dataset was created to assist people in working with model cards. In particular. it was created to support research in the area of model cards and their use. It is possible to use the Hugging Face Hub API or client library to download model cards and this option may be preferable if you have a very specific use case or require a different format.
### Source Data
The source data is 'URL' files for models hosted on the Hugging Face Hub. We do not include any other supplementary files that may be included in the model card directory.
#### Data Collection and Processing
The data is downloaded using a CRON job on a daily basis.
#### Who are the source data producers?
The source data producers are the creators of the model cards on the Hugging Face Hub. This includes a broad variety of people from the community ranging from large companies to individual researchers. We do not gather any information about who created the model card in this repository although this information can be gathered from the Hugging Face Hub API.
### Annotations [optional]
There are no additional annotations in this dataset beyond the model card content.
#### Annotation process
N/A
#### Who are the annotators?
N/A
#### Personal and Sensitive Information
We make no effort to anonymize the data. Whilst we don't expect the majority of model cards to contain personal or sensitive information, it is possible that some model cards may contain this information. Model cards may also link to websites or email addresses.
## Bias, Risks, and Limitations
Model cards are created by the community and we do not have any control over the content of the model cards. We do not review the content of the model cards and we do not make any claims about the accuracy of the information in the model cards.
Some model cards will themselves discuss bias and sometimes this is done by providing examples of bias in either the training data or the responses provided by the model. As a result this dataset may contain examples of bias.
Whilst we do not directly download any images linked to the model cards, some model cards may include images. Some of these images may not be suitable for all audiences.
### Recommendations
No formal citation is required for this dataset but if you use this dataset in your work, please include a link to this dataset page.
## Dataset Card Authors
@davanstrien
## Dataset Card Contact
@davanstrien | [
"# Dataset Card for Hugging Face Hub Model Cards with Embeddings\n\nThis dataset consists of model cards for models hosted on the Hugging Face Hub. The model cards are created by the community and provide information about the model, its performance, its intended uses, and more. \nThis dataset is updated on a daily basis and includes publicly available models on the Hugging Face Hub.\n\nThis dataset is made available to help support users wanting to work with a large number of Model Cards from the Hub. We hope that this dataset will help support research in the area of Model Cards and their use but the format of this dataset may not be useful for all use cases. If there are other features that you would like to see included in this dataset, please open a new discussion. \n\nThis dataset is the same as the Hugging Face Hub Model Cards dataset but with the addition of embeddings for each model card. The embeddings are generated using the jinaai/jina-embeddings-v2-base-en model.",
"## Dataset Details",
"### Dataset Description\n\n\n- Curated by: Daniel van Strien\n- Language(s) (NLP): Model cards on the Hugging Face Hub are predominantly in English but may include other languages.",
"## Uses\n\nThere are a number of potential uses for this dataset including:\n- text mining to find common themes in model cards\n- analysis of the model card format/content\n- topic modelling of model cards\n- analysis of the model card metadata\n- training language models on model cards\n- build a recommender system for model cards\n- build a search engine for model cards",
"### Out-of-Scope Use",
"## Dataset Structure\n\nThis dataset has a single split.",
"## Dataset Creation",
"### Curation Rationale\n\n\n\nThe dataset was created to assist people in working with model cards. In particular. it was created to support research in the area of model cards and their use. It is possible to use the Hugging Face Hub API or client library to download model cards and this option may be preferable if you have a very specific use case or require a different format.",
"### Source Data\n\nThe source data is 'URL' files for models hosted on the Hugging Face Hub. We do not include any other supplementary files that may be included in the model card directory.",
"#### Data Collection and Processing\n\n\n\nThe data is downloaded using a CRON job on a daily basis.",
"#### Who are the source data producers?\n\nThe source data producers are the creators of the model cards on the Hugging Face Hub. This includes a broad variety of people from the community ranging from large companies to individual researchers. We do not gather any information about who created the model card in this repository although this information can be gathered from the Hugging Face Hub API.",
"### Annotations [optional]\n\nThere are no additional annotations in this dataset beyond the model card content.",
"#### Annotation process\n\nN/A",
"#### Who are the annotators?\n\n\n\nN/A",
"#### Personal and Sensitive Information\n\n\n\nWe make no effort to anonymize the data. Whilst we don't expect the majority of model cards to contain personal or sensitive information, it is possible that some model cards may contain this information. Model cards may also link to websites or email addresses.",
"## Bias, Risks, and Limitations\n\n\n\nModel cards are created by the community and we do not have any control over the content of the model cards. We do not review the content of the model cards and we do not make any claims about the accuracy of the information in the model cards. \nSome model cards will themselves discuss bias and sometimes this is done by providing examples of bias in either the training data or the responses provided by the model. As a result this dataset may contain examples of bias. \n\nWhilst we do not directly download any images linked to the model cards, some model cards may include images. Some of these images may not be suitable for all audiences.",
"### Recommendations\n\n\n\n\nNo formal citation is required for this dataset but if you use this dataset in your work, please include a link to this dataset page.",
"## Dataset Card Authors \n\n@davanstrien",
"## Dataset Card Contact\n\n@davanstrien"
] | [
"TAGS\n#task_categories-text-retrieval #size_categories-100K<n<1M #region-us \n",
"# Dataset Card for Hugging Face Hub Model Cards with Embeddings\n\nThis dataset consists of model cards for models hosted on the Hugging Face Hub. The model cards are created by the community and provide information about the model, its performance, its intended uses, and more. \nThis dataset is updated on a daily basis and includes publicly available models on the Hugging Face Hub.\n\nThis dataset is made available to help support users wanting to work with a large number of Model Cards from the Hub. We hope that this dataset will help support research in the area of Model Cards and their use but the format of this dataset may not be useful for all use cases. If there are other features that you would like to see included in this dataset, please open a new discussion. \n\nThis dataset is the same as the Hugging Face Hub Model Cards dataset but with the addition of embeddings for each model card. The embeddings are generated using the jinaai/jina-embeddings-v2-base-en model.",
"## Dataset Details",
"### Dataset Description\n\n\n- Curated by: Daniel van Strien\n- Language(s) (NLP): Model cards on the Hugging Face Hub are predominantly in English but may include other languages.",
"## Uses\n\nThere are a number of potential uses for this dataset including:\n- text mining to find common themes in model cards\n- analysis of the model card format/content\n- topic modelling of model cards\n- analysis of the model card metadata\n- training language models on model cards\n- build a recommender system for model cards\n- build a search engine for model cards",
"### Out-of-Scope Use",
"## Dataset Structure\n\nThis dataset has a single split.",
"## Dataset Creation",
"### Curation Rationale\n\n\n\nThe dataset was created to assist people in working with model cards. In particular. it was created to support research in the area of model cards and their use. It is possible to use the Hugging Face Hub API or client library to download model cards and this option may be preferable if you have a very specific use case or require a different format.",
"### Source Data\n\nThe source data is 'URL' files for models hosted on the Hugging Face Hub. We do not include any other supplementary files that may be included in the model card directory.",
"#### Data Collection and Processing\n\n\n\nThe data is downloaded using a CRON job on a daily basis.",
"#### Who are the source data producers?\n\nThe source data producers are the creators of the model cards on the Hugging Face Hub. This includes a broad variety of people from the community ranging from large companies to individual researchers. We do not gather any information about who created the model card in this repository although this information can be gathered from the Hugging Face Hub API.",
"### Annotations [optional]\n\nThere are no additional annotations in this dataset beyond the model card content.",
"#### Annotation process\n\nN/A",
"#### Who are the annotators?\n\n\n\nN/A",
"#### Personal and Sensitive Information\n\n\n\nWe make no effort to anonymize the data. Whilst we don't expect the majority of model cards to contain personal or sensitive information, it is possible that some model cards may contain this information. Model cards may also link to websites or email addresses.",
"## Bias, Risks, and Limitations\n\n\n\nModel cards are created by the community and we do not have any control over the content of the model cards. We do not review the content of the model cards and we do not make any claims about the accuracy of the information in the model cards. \nSome model cards will themselves discuss bias and sometimes this is done by providing examples of bias in either the training data or the responses provided by the model. As a result this dataset may contain examples of bias. \n\nWhilst we do not directly download any images linked to the model cards, some model cards may include images. Some of these images may not be suitable for all audiences.",
"### Recommendations\n\n\n\n\nNo formal citation is required for this dataset but if you use this dataset in your work, please include a link to this dataset page.",
"## Dataset Card Authors \n\n@davanstrien",
"## Dataset Card Contact\n\n@davanstrien"
] | [
30,
226,
4,
44,
77,
9,
14,
5,
80,
43,
22,
84,
26,
8,
12,
62,
147,
37,
11,
10
] | [
"passage: TAGS\n#task_categories-text-retrieval #size_categories-100K<n<1M #region-us \n# Dataset Card for Hugging Face Hub Model Cards with Embeddings\n\nThis dataset consists of model cards for models hosted on the Hugging Face Hub. The model cards are created by the community and provide information about the model, its performance, its intended uses, and more. \nThis dataset is updated on a daily basis and includes publicly available models on the Hugging Face Hub.\n\nThis dataset is made available to help support users wanting to work with a large number of Model Cards from the Hub. We hope that this dataset will help support research in the area of Model Cards and their use but the format of this dataset may not be useful for all use cases. If there are other features that you would like to see included in this dataset, please open a new discussion. \n\nThis dataset is the same as the Hugging Face Hub Model Cards dataset but with the addition of embeddings for each model card. The embeddings are generated using the jinaai/jina-embeddings-v2-base-en model.## Dataset Details### Dataset Description\n\n\n- Curated by: Daniel van Strien\n- Language(s) (NLP): Model cards on the Hugging Face Hub are predominantly in English but may include other languages.## Uses\n\nThere are a number of potential uses for this dataset including:\n- text mining to find common themes in model cards\n- analysis of the model card format/content\n- topic modelling of model cards\n- analysis of the model card metadata\n- training language models on model cards\n- build a recommender system for model cards\n- build a search engine for model cards### Out-of-Scope Use## Dataset Structure\n\nThis dataset has a single split.## Dataset Creation### Curation Rationale\n\n\n\nThe dataset was created to assist people in working with model cards. In particular. it was created to support research in the area of model cards and their use. It is possible to use the Hugging Face Hub API or client library to download model cards and this option may be preferable if you have a very specific use case or require a different format."
] |
53fe4ea127b2c6921d8c09306933f4ff52dba1d4 | # [doc] image dataset 7
This dataset contains 2 jpeg files in the `red` directory and 2 jpeg files in the `green` directory. | datasets-examples/doc-image-7 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T18:43:07+00:00 | {"size_categories": ["n<1K"]} | 2023-11-30T19:33:26+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 7
This dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory. | [
"# [doc] image dataset 7\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 7\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory."
] | [
16,
37
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 7\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory."
] |
3fd60c5b7b1ba1a4186edd88b078a4ad20647949 | # [doc] image dataset 6
This dataset contains 4 jpeg files in the `train/images/` subdirectory, along with a `train/metadata.csv` file that provides the data for other columns. The metadata file contains relative paths to the images. | datasets-examples/doc-image-6 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T18:43:40+00:00 | {"size_categories": ["n<1K"]} | 2024-01-05T18:19:12+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 6
This dataset contains 4 jpeg files in the 'train/images/' subdirectory, along with a 'train/URL' file that provides the data for other columns. The metadata file contains relative paths to the images. | [
"# [doc] image dataset 6\n\nThis dataset contains 4 jpeg files in the 'train/images/' subdirectory, along with a 'train/URL' file that provides the data for other columns. The metadata file contains relative paths to the images."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 6\n\nThis dataset contains 4 jpeg files in the 'train/images/' subdirectory, along with a 'train/URL' file that provides the data for other columns. The metadata file contains relative paths to the images."
] | [
16,
64
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 6\n\nThis dataset contains 4 jpeg files in the 'train/images/' subdirectory, along with a 'train/URL' file that provides the data for other columns. The metadata file contains relative paths to the images."
] |
96b4a56c8073dcc8b6e36b5b631dfad60426334a |
# Dataset Summary
This dataset contains all MQM human annotations from previous [WMT Metrics shared tasks](https://wmt-metrics-task.github.io/) and the MQM annotations from [Experts, Errors, and Context](https://aclanthology.org/2021.tacl-1.87/) in a form of error spans. Moreover, it contains some hallucinations used in the training of [XCOMET models](https://huggingface.co/Unbabel/XCOMET-XXL).
**Please note that this is not an official release of the data** and the original data can be found [here](https://github.com/google/wmt-mqm-human-evaluation).
The data is organised into 8 columns:
- src: input text
- mt: translation
- ref: reference translation
- annotations: List of error spans (dictionaries with 'start', 'end', 'severity', 'text')
- lp: language pair
While `en-ru` was annotated by Unbabel, `en-de` and `zh-en` was annotated by Google. This means that for en-de and zh-en you will only find minor and major errors while for en-ru you can find a few critical errors.
## Python usage:
```python
from datasets import load_dataset
dataset = load_dataset("RicardoRei/wmt-mqm-error-spans", split="train")
```
There is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :
```python
# split by LP
data = dataset.filter(lambda example: example["lp"] == "en-de")
```
## Citation Information
If you use this data please cite the following works:
- [Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation](https://aclanthology.org/2021.tacl-1.87/)
- [Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain](https://aclanthology.org/2021.wmt-1.73/)
- [Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust](https://aclanthology.org/2022.wmt-1.2/)
- [xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection](https://arxiv.org/pdf/2310.10482.pdf)
| RicardoRei/wmt-mqm-error-spans | [
"size_categories:100K<n<1M",
"language:en",
"language:de",
"language:ru",
"language:zh",
"license:apache-2.0",
"mt-evaluation",
"WMT",
"MQM",
"arxiv:2310.10482",
"region:us"
] | 2023-11-30T18:55:52+00:00 | {"language": ["en", "de", "ru", "zh"], "license": "apache-2.0", "size_categories": ["100K<n<1M"], "tags": ["mt-evaluation", "WMT", "MQM"]} | 2023-11-30T19:14:18+00:00 | [
"2310.10482"
] | [
"en",
"de",
"ru",
"zh"
] | TAGS
#size_categories-100K<n<1M #language-English #language-German #language-Russian #language-Chinese #license-apache-2.0 #mt-evaluation #WMT #MQM #arxiv-2310.10482 #region-us
|
# Dataset Summary
This dataset contains all MQM human annotations from previous WMT Metrics shared tasks and the MQM annotations from Experts, Errors, and Context in a form of error spans. Moreover, it contains some hallucinations used in the training of XCOMET models.
Please note that this is not an official release of the data and the original data can be found here.
The data is organised into 8 columns:
- src: input text
- mt: translation
- ref: reference translation
- annotations: List of error spans (dictionaries with 'start', 'end', 'severity', 'text')
- lp: language pair
While 'en-ru' was annotated by Unbabel, 'en-de' and 'zh-en' was annotated by Google. This means that for en-de and zh-en you will only find minor and major errors while for en-ru you can find a few critical errors.
## Python usage:
There is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :
If you use this data please cite the following works:
- Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation
- Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain
- Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust
- xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection
| [
"# Dataset Summary\n\nThis dataset contains all MQM human annotations from previous WMT Metrics shared tasks and the MQM annotations from Experts, Errors, and Context in a form of error spans. Moreover, it contains some hallucinations used in the training of XCOMET models.\n\nPlease note that this is not an official release of the data and the original data can be found here.\n\nThe data is organised into 8 columns:\n\n- src: input text\n- mt: translation\n- ref: reference translation\n- annotations: List of error spans (dictionaries with 'start', 'end', 'severity', 'text')\n- lp: language pair\n \n\nWhile 'en-ru' was annotated by Unbabel, 'en-de' and 'zh-en' was annotated by Google. This means that for en-de and zh-en you will only find minor and major errors while for en-ru you can find a few critical errors.",
"## Python usage:\n\n\n\nThere is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :\n\n\n\n\n\nIf you use this data please cite the following works:\n- Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation\n- Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain\n- Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust\n- xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection"
] | [
"TAGS\n#size_categories-100K<n<1M #language-English #language-German #language-Russian #language-Chinese #license-apache-2.0 #mt-evaluation #WMT #MQM #arxiv-2310.10482 #region-us \n",
"# Dataset Summary\n\nThis dataset contains all MQM human annotations from previous WMT Metrics shared tasks and the MQM annotations from Experts, Errors, and Context in a form of error spans. Moreover, it contains some hallucinations used in the training of XCOMET models.\n\nPlease note that this is not an official release of the data and the original data can be found here.\n\nThe data is organised into 8 columns:\n\n- src: input text\n- mt: translation\n- ref: reference translation\n- annotations: List of error spans (dictionaries with 'start', 'end', 'severity', 'text')\n- lp: language pair\n \n\nWhile 'en-ru' was annotated by Unbabel, 'en-de' and 'zh-en' was annotated by Google. This means that for en-de and zh-en you will only find minor and major errors while for en-ru you can find a few critical errors.",
"## Python usage:\n\n\n\nThere is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :\n\n\n\n\n\nIf you use this data please cite the following works:\n- Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation\n- Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain\n- Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust\n- xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection"
] | [
65,
226,
154
] | [
"passage: TAGS\n#size_categories-100K<n<1M #language-English #language-German #language-Russian #language-Chinese #license-apache-2.0 #mt-evaluation #WMT #MQM #arxiv-2310.10482 #region-us \n# Dataset Summary\n\nThis dataset contains all MQM human annotations from previous WMT Metrics shared tasks and the MQM annotations from Experts, Errors, and Context in a form of error spans. Moreover, it contains some hallucinations used in the training of XCOMET models.\n\nPlease note that this is not an official release of the data and the original data can be found here.\n\nThe data is organised into 8 columns:\n\n- src: input text\n- mt: translation\n- ref: reference translation\n- annotations: List of error spans (dictionaries with 'start', 'end', 'severity', 'text')\n- lp: language pair\n \n\nWhile 'en-ru' was annotated by Unbabel, 'en-de' and 'zh-en' was annotated by Google. This means that for en-de and zh-en you will only find minor and major errors while for en-ru you can find a few critical errors.## Python usage:\n\n\n\nThere is no standard train/test split for this dataset but you can easily split it according to year, language pair or domain. E.g. :\n\n\n\n\n\nIf you use this data please cite the following works:\n- Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation\n- Results of the WMT21 Metrics Shared Task: Evaluating Metrics with Expert-based Human Evaluations on TED and News Domain\n- Results of WMT22 Metrics Shared Task: Stop Using BLEU – Neural Metrics Are Better and More Robust\n- xCOMET: Transparent Machine Translation Evaluation through Fine-grained Error Detection"
] |
c6a1667cbee43be6e4322f3c06a16af7fb662ef8 | # Dataset Card for "emrqa_train"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | hippocrates/emrqa_train | [
"region:us"
] | 2023-11-30T19:03:35+00:00 | {"dataset_info": {"features": [{"name": "id", "dtype": "string"}, {"name": "conversations", "list": [{"name": "from", "dtype": "string"}, {"name": "value", "dtype": "string"}]}, {"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 156916965, "num_examples": 152907}, {"name": "valid", "num_bytes": 156916965, "num_examples": 152907}, {"name": "test", "num_bytes": 26945615, "num_examples": 26804}], "download_size": 89918618, "dataset_size": 340779545}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "valid", "path": "data/valid-*"}, {"split": "test", "path": "data/test-*"}]}]} | 2024-01-07T04:19:55+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "emrqa_train"
More Information needed | [
"# Dataset Card for \"emrqa_train\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"emrqa_train\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"emrqa_train\"\n\nMore Information needed"
] |
a6aae301b9f0d102832b388617fe3480bad80bd8 | # MiniMuSiQue by Morph Labs
**https://morph.so/blog/self-teaching/**
We describe two evaluation datasets that we have derived from the MuSiQue multi-hop question-answering dataset, called MiniMuSiQue-hard (filtered for questions answerable by GPT-4 but not GPT-3.5, where performance significantly degrades if the first pivot document is removed) and MiniMuSiQue-easy (a larger dataset of convoluted off-distribution single-hop question-answer pairs).
## Table of Contents
1. **<a href="https://huggingface.co/morph-labs/MiniMuSiQue#dataset-description" target="_blank">Dataset Description</a>**
2. **<a href="https://huggingface.co/morph-labs/MiniMuSiQue#uses" target="_blank">Uses</a>**
3. **<a href="https://huggingface.co/morph-labs/MiniMuSiQue#contact" target="_blank">Contact</a>**
4. **<a href="https://huggingface.co/morph-labs/MiniMuSiQue#blogpost-and-citation" target="_blank">Blogpost and Citation</a>**
### Dataset Description
We refined the MuSiQue dataset to focus on questions that demand complex multi-hop reasoning, by selecting questions which (1) GPT-4 could answer but GPT-3.5 could not, and which (2) were not answerable without the context relevant to the first reasoning step (the "first hop pivot document") for each question. Specifically, we selected 768 random examples from the MuSiQue training set, ranked them based on a combined score of difficulty (measured by the difference in ROUGE-L recall between GPT-4 and GPT-3.5) and the necessity for multi-hop reasoning (assessed by the change in ROUGE-L recall when the first hop pivot document was removed). We refer to the top-ranked 128 examples as MiniMuSiQue, and obtain MiniMuSiQue-hard by associating the original difficult MuSiQue multi-hop question-answer pair to each example. To additionally test off-distribution single-hop factual recall, for each example we synthesized convoluted off-distribution single-hop question-answer pairs for up to five entities per document in MiniMuSiQue, resulting in the much larger single-hop dataset MiniMuSiQue-easy. Each MiniMuSiQue example consists of twenty documents sampled from different Wikipedia articles, to which we associate a hard MuSiQue multi-hop reasoning question for MiniMuSiQue, and many single-hop questions for MiniMuSiQue-easy.
- **Developed by:** **<a href="https://www.morph.so" target="_blank">Morph Labs</a>**
- **Refined from:** **<a href="https://arxiv.org/abs/2108.00573" target="_blank">MuSiQue</a>**
- **Language(s):** English
- **License:** **<a href="https://www.apache.org/licenses/LICENSE-2.0" target="_blank">Apache 2.0</a>**
## Uses
A particularly challenging form of question for models historically has been multi-hop questions, which require a series of interconnected reasoning steps over multiple documents. However, creating multi-hop questions that truly necessitate knowledge-based reasoning is challenging. For instance, early benchmarks like HotpotQA were found to be largely solvable through shortcuts. The construction of questions and corresponding contexts that avoid such shortcuts, and verifying their effectiveness, requires a comprehensive dataset development process. The MuSiQue dataset addresses many weaknesses of prior work and contains difficult multi-hop questions less susceptible to shortcuts. We derive MiniMuSiQue from the original MuSiQue to better assess model capabilities to answer multi-hop questions that truly necessitate knowledge-based reasoning.
## Contact
[email protected]
## Blogpost and Citation
**https://morph.so/blog/self-teaching/**
@misc{MiniMuSiQue,
title={MiniMuSiQue},
author={Morph Labs, Jesse Michael Han, Eric Yu, Bentley Long, Pranav Mital, Brando Miranda},
year={2023}} | morph-labs/MiniMuSiQue | [
"language:en",
"license:apache-2.0",
"arxiv:2108.00573",
"region:us"
] | 2023-11-30T19:05:12+00:00 | {"language": ["en"], "license": "apache-2.0"} | 2023-12-05T21:14:24+00:00 | [
"2108.00573"
] | [
"en"
] | TAGS
#language-English #license-apache-2.0 #arxiv-2108.00573 #region-us
| # MiniMuSiQue by Morph Labs
!banner
URL
We describe two evaluation datasets that we have derived from the MuSiQue multi-hop question-answering dataset, called MiniMuSiQue-hard (filtered for questions answerable by GPT-4 but not GPT-3.5, where performance significantly degrades if the first pivot document is removed) and MiniMuSiQue-easy (a larger dataset of convoluted off-distribution single-hop question-answer pairs).
## Table of Contents
1. <a href="URL target="_blank">Dataset Description</a>
2. <a href="URL target="_blank">Uses</a>
3. <a href="URL target="_blank">Contact</a>
4. <a href="URL target="_blank">Blogpost and Citation</a>
### Dataset Description
We refined the MuSiQue dataset to focus on questions that demand complex multi-hop reasoning, by selecting questions which (1) GPT-4 could answer but GPT-3.5 could not, and which (2) were not answerable without the context relevant to the first reasoning step (the "first hop pivot document") for each question. Specifically, we selected 768 random examples from the MuSiQue training set, ranked them based on a combined score of difficulty (measured by the difference in ROUGE-L recall between GPT-4 and GPT-3.5) and the necessity for multi-hop reasoning (assessed by the change in ROUGE-L recall when the first hop pivot document was removed). We refer to the top-ranked 128 examples as MiniMuSiQue, and obtain MiniMuSiQue-hard by associating the original difficult MuSiQue multi-hop question-answer pair to each example. To additionally test off-distribution single-hop factual recall, for each example we synthesized convoluted off-distribution single-hop question-answer pairs for up to five entities per document in MiniMuSiQue, resulting in the much larger single-hop dataset MiniMuSiQue-easy. Each MiniMuSiQue example consists of twenty documents sampled from different Wikipedia articles, to which we associate a hard MuSiQue multi-hop reasoning question for MiniMuSiQue, and many single-hop questions for MiniMuSiQue-easy.
- Developed by: <a href="URL" target="_blank">Morph Labs</a>
- Refined from: <a href="URL target="_blank">MuSiQue</a>
- Language(s): English
- License: <a href="URL target="_blank">Apache 2.0</a>
## Uses
A particularly challenging form of question for models historically has been multi-hop questions, which require a series of interconnected reasoning steps over multiple documents. However, creating multi-hop questions that truly necessitate knowledge-based reasoning is challenging. For instance, early benchmarks like HotpotQA were found to be largely solvable through shortcuts. The construction of questions and corresponding contexts that avoid such shortcuts, and verifying their effectiveness, requires a comprehensive dataset development process. The MuSiQue dataset addresses many weaknesses of prior work and contains difficult multi-hop questions less susceptible to shortcuts. We derive MiniMuSiQue from the original MuSiQue to better assess model capabilities to answer multi-hop questions that truly necessitate knowledge-based reasoning.
## Contact
hello@URL
## Blogpost and Citation
URL
@misc{MiniMuSiQue,
title={MiniMuSiQue},
author={Morph Labs, Jesse Michael Han, Eric Yu, Bentley Long, Pranav Mital, Brando Miranda},
year={2023}} | [
"# MiniMuSiQue by Morph Labs\n\n!banner\n\nURL\n\nWe describe two evaluation datasets that we have derived from the MuSiQue multi-hop question-answering dataset, called MiniMuSiQue-hard (filtered for questions answerable by GPT-4 but not GPT-3.5, where performance significantly degrades if the first pivot document is removed) and MiniMuSiQue-easy (a larger dataset of convoluted off-distribution single-hop question-answer pairs).",
"## Table of Contents\n\n\n1. <a href=\"URL target=\"_blank\">Dataset Description</a>\n2. <a href=\"URL target=\"_blank\">Uses</a>\n3. <a href=\"URL target=\"_blank\">Contact</a>\n4. <a href=\"URL target=\"_blank\">Blogpost and Citation</a>",
"### Dataset Description\n\nWe refined the MuSiQue dataset to focus on questions that demand complex multi-hop reasoning, by selecting questions which (1) GPT-4 could answer but GPT-3.5 could not, and which (2) were not answerable without the context relevant to the first reasoning step (the \"first hop pivot document\") for each question. Specifically, we selected 768 random examples from the MuSiQue training set, ranked them based on a combined score of difficulty (measured by the difference in ROUGE-L recall between GPT-4 and GPT-3.5) and the necessity for multi-hop reasoning (assessed by the change in ROUGE-L recall when the first hop pivot document was removed). We refer to the top-ranked 128 examples as MiniMuSiQue, and obtain MiniMuSiQue-hard by associating the original difficult MuSiQue multi-hop question-answer pair to each example. To additionally test off-distribution single-hop factual recall, for each example we synthesized convoluted off-distribution single-hop question-answer pairs for up to five entities per document in MiniMuSiQue, resulting in the much larger single-hop dataset MiniMuSiQue-easy. Each MiniMuSiQue example consists of twenty documents sampled from different Wikipedia articles, to which we associate a hard MuSiQue multi-hop reasoning question for MiniMuSiQue, and many single-hop questions for MiniMuSiQue-easy.\n\n\n- Developed by: <a href=\"URL\" target=\"_blank\">Morph Labs</a>\n- Refined from: <a href=\"URL target=\"_blank\">MuSiQue</a>\n- Language(s): English\n- License: <a href=\"URL target=\"_blank\">Apache 2.0</a>",
"## Uses\n\nA particularly challenging form of question for models historically has been multi-hop questions, which require a series of interconnected reasoning steps over multiple documents. However, creating multi-hop questions that truly necessitate knowledge-based reasoning is challenging. For instance, early benchmarks like HotpotQA were found to be largely solvable through shortcuts. The construction of questions and corresponding contexts that avoid such shortcuts, and verifying their effectiveness, requires a comprehensive dataset development process. The MuSiQue dataset addresses many weaknesses of prior work and contains difficult multi-hop questions less susceptible to shortcuts. We derive MiniMuSiQue from the original MuSiQue to better assess model capabilities to answer multi-hop questions that truly necessitate knowledge-based reasoning.",
"## Contact\n\nhello@URL",
"## Blogpost and Citation\n\nURL\n\n@misc{MiniMuSiQue,\n title={MiniMuSiQue}, \n author={Morph Labs, Jesse Michael Han, Eric Yu, Bentley Long, Pranav Mital, Brando Miranda},\n year={2023}}"
] | [
"TAGS\n#language-English #license-apache-2.0 #arxiv-2108.00573 #region-us \n",
"# MiniMuSiQue by Morph Labs\n\n!banner\n\nURL\n\nWe describe two evaluation datasets that we have derived from the MuSiQue multi-hop question-answering dataset, called MiniMuSiQue-hard (filtered for questions answerable by GPT-4 but not GPT-3.5, where performance significantly degrades if the first pivot document is removed) and MiniMuSiQue-easy (a larger dataset of convoluted off-distribution single-hop question-answer pairs).",
"## Table of Contents\n\n\n1. <a href=\"URL target=\"_blank\">Dataset Description</a>\n2. <a href=\"URL target=\"_blank\">Uses</a>\n3. <a href=\"URL target=\"_blank\">Contact</a>\n4. <a href=\"URL target=\"_blank\">Blogpost and Citation</a>",
"### Dataset Description\n\nWe refined the MuSiQue dataset to focus on questions that demand complex multi-hop reasoning, by selecting questions which (1) GPT-4 could answer but GPT-3.5 could not, and which (2) were not answerable without the context relevant to the first reasoning step (the \"first hop pivot document\") for each question. Specifically, we selected 768 random examples from the MuSiQue training set, ranked them based on a combined score of difficulty (measured by the difference in ROUGE-L recall between GPT-4 and GPT-3.5) and the necessity for multi-hop reasoning (assessed by the change in ROUGE-L recall when the first hop pivot document was removed). We refer to the top-ranked 128 examples as MiniMuSiQue, and obtain MiniMuSiQue-hard by associating the original difficult MuSiQue multi-hop question-answer pair to each example. To additionally test off-distribution single-hop factual recall, for each example we synthesized convoluted off-distribution single-hop question-answer pairs for up to five entities per document in MiniMuSiQue, resulting in the much larger single-hop dataset MiniMuSiQue-easy. Each MiniMuSiQue example consists of twenty documents sampled from different Wikipedia articles, to which we associate a hard MuSiQue multi-hop reasoning question for MiniMuSiQue, and many single-hop questions for MiniMuSiQue-easy.\n\n\n- Developed by: <a href=\"URL\" target=\"_blank\">Morph Labs</a>\n- Refined from: <a href=\"URL target=\"_blank\">MuSiQue</a>\n- Language(s): English\n- License: <a href=\"URL target=\"_blank\">Apache 2.0</a>",
"## Uses\n\nA particularly challenging form of question for models historically has been multi-hop questions, which require a series of interconnected reasoning steps over multiple documents. However, creating multi-hop questions that truly necessitate knowledge-based reasoning is challenging. For instance, early benchmarks like HotpotQA were found to be largely solvable through shortcuts. The construction of questions and corresponding contexts that avoid such shortcuts, and verifying their effectiveness, requires a comprehensive dataset development process. The MuSiQue dataset addresses many weaknesses of prior work and contains difficult multi-hop questions less susceptible to shortcuts. We derive MiniMuSiQue from the original MuSiQue to better assess model capabilities to answer multi-hop questions that truly necessitate knowledge-based reasoning.",
"## Contact\n\nhello@URL",
"## Blogpost and Citation\n\nURL\n\n@misc{MiniMuSiQue,\n title={MiniMuSiQue}, \n author={Morph Labs, Jesse Michael Han, Eric Yu, Bentley Long, Pranav Mital, Brando Miranda},\n year={2023}}"
] | [
26,
113,
76,
414,
175,
6,
61
] | [
"passage: TAGS\n#language-English #license-apache-2.0 #arxiv-2108.00573 #region-us \n# MiniMuSiQue by Morph Labs\n\n!banner\n\nURL\n\nWe describe two evaluation datasets that we have derived from the MuSiQue multi-hop question-answering dataset, called MiniMuSiQue-hard (filtered for questions answerable by GPT-4 but not GPT-3.5, where performance significantly degrades if the first pivot document is removed) and MiniMuSiQue-easy (a larger dataset of convoluted off-distribution single-hop question-answer pairs).## Table of Contents\n\n\n1. <a href=\"URL target=\"_blank\">Dataset Description</a>\n2. <a href=\"URL target=\"_blank\">Uses</a>\n3. <a href=\"URL target=\"_blank\">Contact</a>\n4. <a href=\"URL target=\"_blank\">Blogpost and Citation</a>"
] |
800ab9c2a052a56a7ecd796e26c888ab8dce6e03 |
# Bangumi Image Base of Full Metal Panic Fumoffu
This is the image base of bangumi Full Metal Panic Fumoffu, we detected 22 characters, 1168 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 333 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 43 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 238 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 26 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 35 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 29 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 21 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 92 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 13 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 14 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 16 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 33 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 18 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 10 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 12 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 8 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 11 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 20 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 49 | [Download](18/dataset.zip) |  |  |  |  |  |  |  |  |
| 19 | 7 | [Download](19/dataset.zip) |  |  |  |  |  |  |  | N/A |
| 20 | 19 | [Download](20/dataset.zip) |  |  |  |  |  |  |  |  |
| noise | 121 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/fullmetalpanicfumoffu | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T19:22:16+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T19:56:58+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of Full Metal Panic Fumoffu
==============================================
This is the image base of bangumi Full Metal Panic Fumoffu, we detected 22 characters, 1168 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
a7012e6a58dbb8bf51844400013645191f47ec92 |
# Bangumi Image Base of The Twelve Kingdoms
This is the image base of bangumi The Twelve Kingdoms, we detected 62 characters, 4697 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 35 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 731 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 72 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 20 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 88 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 399 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 254 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 163 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 466 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 80 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 82 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 72 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 31 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 35 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 42 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 53 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 22 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 153 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 107 | [Download](18/dataset.zip) |  |  |  |  |  |  |  |  |
| 19 | 93 | [Download](19/dataset.zip) |  |  |  |  |  |  |  |  |
| 20 | 70 | [Download](20/dataset.zip) |  |  |  |  |  |  |  |  |
| 21 | 111 | [Download](21/dataset.zip) |  |  |  |  |  |  |  |  |
| 22 | 60 | [Download](22/dataset.zip) |  |  |  |  |  |  |  |  |
| 23 | 14 | [Download](23/dataset.zip) |  |  |  |  |  |  |  |  |
| 24 | 25 | [Download](24/dataset.zip) |  |  |  |  |  |  |  |  |
| 25 | 70 | [Download](25/dataset.zip) |  |  |  |  |  |  |  |  |
| 26 | 32 | [Download](26/dataset.zip) |  |  |  |  |  |  |  |  |
| 27 | 121 | [Download](27/dataset.zip) |  |  |  |  |  |  |  |  |
| 28 | 15 | [Download](28/dataset.zip) |  |  |  |  |  |  |  |  |
| 29 | 24 | [Download](29/dataset.zip) |  |  |  |  |  |  |  |  |
| 30 | 17 | [Download](30/dataset.zip) |  |  |  |  |  |  |  |  |
| 31 | 16 | [Download](31/dataset.zip) |  |  |  |  |  |  |  |  |
| 32 | 18 | [Download](32/dataset.zip) |  |  |  |  |  |  |  |  |
| 33 | 55 | [Download](33/dataset.zip) |  |  |  |  |  |  |  |  |
| 34 | 16 | [Download](34/dataset.zip) |  |  |  |  |  |  |  |  |
| 35 | 23 | [Download](35/dataset.zip) |  |  |  |  |  |  |  |  |
| 36 | 44 | [Download](36/dataset.zip) |  |  |  |  |  |  |  |  |
| 37 | 11 | [Download](37/dataset.zip) |  |  |  |  |  |  |  |  |
| 38 | 15 | [Download](38/dataset.zip) |  |  |  |  |  |  |  |  |
| 39 | 23 | [Download](39/dataset.zip) |  |  |  |  |  |  |  |  |
| 40 | 9 | [Download](40/dataset.zip) |  |  |  |  |  |  |  |  |
| 41 | 22 | [Download](41/dataset.zip) |  |  |  |  |  |  |  |  |
| 42 | 149 | [Download](42/dataset.zip) |  |  |  |  |  |  |  |  |
| 43 | 134 | [Download](43/dataset.zip) |  |  |  |  |  |  |  |  |
| 44 | 27 | [Download](44/dataset.zip) |  |  |  |  |  |  |  |  |
| 45 | 42 | [Download](45/dataset.zip) |  |  |  |  |  |  |  |  |
| 46 | 11 | [Download](46/dataset.zip) |  |  |  |  |  |  |  |  |
| 47 | 18 | [Download](47/dataset.zip) |  |  |  |  |  |  |  |  |
| 48 | 14 | [Download](48/dataset.zip) |  |  |  |  |  |  |  |  |
| 49 | 55 | [Download](49/dataset.zip) |  |  |  |  |  |  |  |  |
| 50 | 27 | [Download](50/dataset.zip) |  |  |  |  |  |  |  |  |
| 51 | 20 | [Download](51/dataset.zip) |  |  |  |  |  |  |  |  |
| 52 | 11 | [Download](52/dataset.zip) |  |  |  |  |  |  |  |  |
| 53 | 55 | [Download](53/dataset.zip) |  |  |  |  |  |  |  |  |
| 54 | 30 | [Download](54/dataset.zip) |  |  |  |  |  |  |  |  |
| 55 | 13 | [Download](55/dataset.zip) |  |  |  |  |  |  |  |  |
| 56 | 26 | [Download](56/dataset.zip) |  |  |  |  |  |  |  |  |
| 57 | 109 | [Download](57/dataset.zip) |  |  |  |  |  |  |  |  |
| 58 | 14 | [Download](58/dataset.zip) |  |  |  |  |  |  |  |  |
| 59 | 20 | [Download](59/dataset.zip) |  |  |  |  |  |  |  |  |
| 60 | 13 | [Download](60/dataset.zip) |  |  |  |  |  |  |  |  |
| noise | 100 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/thetwelvekingdoms | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T19:23:01+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T21:42:54+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of The Twelve Kingdoms
=========================================
This is the image base of bangumi The Twelve Kingdoms, we detected 62 characters, 4697 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
cf9b1bdb421f966b7b6fd372f56fad4b6b805e0f | # Dataset Card for "cai-conversation-prod-h4-harmless"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | vwxyzjn/cai-conversation-prod-h4-harmless | [
"region:us"
] | 2023-11-30T19:30:03+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train_sft", "path": "data/train_sft-*"}, {"split": "test_sft", "path": "data/test_sft-*"}, {"split": "train_prefs", "path": "data/train_prefs-*"}, {"split": "test_prefs", "path": "data/test_prefs-*"}]}], "dataset_info": {"features": [{"name": "index", "dtype": "int64"}, {"name": "prompt", "dtype": "string"}, {"name": "init_prompt", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "init_response", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "critic_prompt", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "critic_response", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "revision_prompt", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "revision_response", "struct": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "messages", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "chosen", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}, {"name": "rejected", "list": [{"name": "content", "dtype": "string"}, {"name": "role", "dtype": "string"}]}], "splits": [{"name": "train_sft", "num_bytes": 80509011.39550994, "num_examples": 21289}, {"name": "test_sft", "num_bytes": 4409523.340505145, "num_examples": 1156}, {"name": "train_prefs", "num_bytes": 80509011.39550994, "num_examples": 21289}, {"name": "test_prefs", "num_bytes": 4413337.807062675, "num_examples": 1157}], "download_size": 52332286, "dataset_size": 169840883.93858773}} | 2023-12-01T14:30:02+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "cai-conversation-prod-h4-harmless"
More Information needed | [
"# Dataset Card for \"cai-conversation-prod-h4-harmless\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"cai-conversation-prod-h4-harmless\"\n\nMore Information needed"
] | [
6,
24
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"cai-conversation-prod-h4-harmless\"\n\nMore Information needed"
] |
217f7ab94e35b877e3192debf0af4cbc72ccdea6 | # [doc] image dataset 8
This dataset contains 4 jpeg files in the `{train,test}/{green,red}/`subdirectories, which specify the splits and the image classes. | datasets-examples/doc-image-8 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T19:35:01+00:00 | {"size_categories": ["n<1K"]} | 2023-11-30T19:39:56+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 8
This dataset contains 4 jpeg files in the '{train,test}/{green,red}/'subdirectories, which specify the splits and the image classes. | [
"# [doc] image dataset 8\n\nThis dataset contains 4 jpeg files in the '{train,test}/{green,red}/'subdirectories, which specify the splits and the image classes."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 8\n\nThis dataset contains 4 jpeg files in the '{train,test}/{green,red}/'subdirectories, which specify the splits and the image classes."
] | [
16,
50
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 8\n\nThis dataset contains 4 jpeg files in the '{train,test}/{green,red}/'subdirectories, which specify the splits and the image classes."
] |
7f070475c9c9e951478b2b537b053f6b167a928d | # [doc] image dataset 10
This dataset contains a parquet file that contains an image column. | datasets-examples/doc-image-10 | [
"size_categories:n<1K",
"region:us"
] | 2023-11-30T19:45:43+00:00 | {"size_categories": ["n<1K"]} | 2023-12-01T17:26:24+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 10
This dataset contains a parquet file that contains an image column. | [
"# [doc] image dataset 10\n\nThis dataset contains a parquet file that contains an image column."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 10\n\nThis dataset contains a parquet file that contains an image column."
] | [
16,
26
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 10\n\nThis dataset contains a parquet file that contains an image column."
] |
d8bae27cc677b2b9b4bef8c2a970937fd003b8cb |
# Bangumi Image Base of Black Lagoon
This is the image base of bangumi Black Lagoon, we detected 24 characters, 2637 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 339 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 101 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 76 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 396 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 82 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 232 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 53 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 68 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 31 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 39 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 36 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 68 | [Download](11/dataset.zip) |  |  |  |  |  |  |  |  |
| 12 | 18 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| 13 | 114 | [Download](13/dataset.zip) |  |  |  |  |  |  |  |  |
| 14 | 97 | [Download](14/dataset.zip) |  |  |  |  |  |  |  |  |
| 15 | 76 | [Download](15/dataset.zip) |  |  |  |  |  |  |  |  |
| 16 | 10 | [Download](16/dataset.zip) |  |  |  |  |  |  |  |  |
| 17 | 39 | [Download](17/dataset.zip) |  |  |  |  |  |  |  |  |
| 18 | 76 | [Download](18/dataset.zip) |  |  |  |  |  |  |  |  |
| 19 | 54 | [Download](19/dataset.zip) |  |  |  |  |  |  |  |  |
| 20 | 15 | [Download](20/dataset.zip) |  |  |  |  |  |  |  |  |
| 21 | 8 | [Download](21/dataset.zip) |  |  |  |  |  |  |  |  |
| 22 | 11 | [Download](22/dataset.zip) |  |  |  |  |  |  |  |  |
| noise | 598 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/blacklagoon | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T20:53:41+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T22:30:28+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of Black Lagoon
==================================
This is the image base of bangumi Black Lagoon, we detected 24 characters, 2637 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
ba4e55d03b92ff60c6be2c652e5a4b53409bfe11 |
# Bangumi Image Base of Azumanga Daioh
This is the image base of bangumi Azumanga Daioh, we detected 14 characters, 3047 images in total. The full dataset is [here](all.zip).
**Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual.** If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| # | Images | Download | Preview 1 | Preview 2 | Preview 3 | Preview 4 | Preview 5 | Preview 6 | Preview 7 | Preview 8 |
|:------|---------:|:---------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|:-------------------------------|
| 0 | 76 | [Download](0/dataset.zip) |  |  |  |  |  |  |  |  |
| 1 | 83 | [Download](1/dataset.zip) |  |  |  |  |  |  |  |  |
| 2 | 607 | [Download](2/dataset.zip) |  |  |  |  |  |  |  |  |
| 3 | 311 | [Download](3/dataset.zip) |  |  |  |  |  |  |  |  |
| 4 | 233 | [Download](4/dataset.zip) |  |  |  |  |  |  |  |  |
| 5 | 502 | [Download](5/dataset.zip) |  |  |  |  |  |  |  |  |
| 6 | 478 | [Download](6/dataset.zip) |  |  |  |  |  |  |  |  |
| 7 | 151 | [Download](7/dataset.zip) |  |  |  |  |  |  |  |  |
| 8 | 31 | [Download](8/dataset.zip) |  |  |  |  |  |  |  |  |
| 9 | 500 | [Download](9/dataset.zip) |  |  |  |  |  |  |  |  |
| 10 | 20 | [Download](10/dataset.zip) |  |  |  |  |  |  |  |  |
| 11 | 7 | [Download](11/dataset.zip) |  |  |  |  |  |  |  | N/A |
| 12 | 10 | [Download](12/dataset.zip) |  |  |  |  |  |  |  |  |
| noise | 38 | [Download](-1/dataset.zip) |  |  |  |  |  |  |  |  |
| BangumiBase/azumangadaioh | [
"size_categories:1K<n<10K",
"license:mit",
"art",
"region:us"
] | 2023-11-30T20:59:44+00:00 | {"license": "mit", "size_categories": ["1K<n<10K"], "tags": ["art"]} | 2023-11-30T22:10:00+00:00 | [] | [] | TAGS
#size_categories-1K<n<10K #license-mit #art #region-us
| Bangumi Image Base of Azumanga Daioh
====================================
This is the image base of bangumi Azumanga Daioh, we detected 14 characters, 3047 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters' preview:
| [] | [
"TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] | [
25
] | [
"passage: TAGS\n#size_categories-1K<n<10K #license-mit #art #region-us \n"
] |
836baaaa6f90f7c19517b599a31ea8c1eb643fd0 | # Dataset Card for "required-deloitte-jobs"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | evkes/required-deloitte-jobs | [
"region:us"
] | 2023-11-30T21:09:38+00:00 | {"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 539333, "num_examples": 327}], "download_size": 205214, "dataset_size": 539333}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-11-30T21:09:40+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "required-deloitte-jobs"
More Information needed | [
"# Dataset Card for \"required-deloitte-jobs\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"required-deloitte-jobs\"\n\nMore Information needed"
] | [
6,
20
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"required-deloitte-jobs\"\n\nMore Information needed"
] |
02e283ab3995ffda87a1c32d231d8ac7fafdd9d5 |
Processed https://huggingface.co/datasets/craigslist_bargains into a canonical text classification dataset.
The text is the dialogue between the buyer and seller. The label is the category of the product which they are discussing.
Note that some records don't have dialogues, i.e., the `text` field is null. I typically handle this by replacing these nulls with the empty string. For example, in pandas:
```python
df["text"] = df["text"].fillna("")
```
For an example of a solution to this task, see the notebook [here](https://github.com/kddubey/cappr/blob/main/demos/huggingface/craigslist_bargains.ipynb).
The notebook which processed the original dataset into this one is available [here](https://github.com/kddubey/stackexchange/blob/main/train_on_test_features/bert/craigslist_bargains.ipynb). | aladar/craigslist_bargains | [
"task_categories:text-classification",
"size_categories:1K<n<10K",
"language:en",
"license:mit",
"region:us"
] | 2023-11-30T21:27:12+00:00 | {"language": ["en"], "license": "mit", "size_categories": ["1K<n<10K"], "task_categories": ["text-classification"]} | 2023-12-27T10:04:14+00:00 | [] | [
"en"
] | TAGS
#task_categories-text-classification #size_categories-1K<n<10K #language-English #license-mit #region-us
|
Processed URL into a canonical text classification dataset.
The text is the dialogue between the buyer and seller. The label is the category of the product which they are discussing.
Note that some records don't have dialogues, i.e., the 'text' field is null. I typically handle this by replacing these nulls with the empty string. For example, in pandas:
For an example of a solution to this task, see the notebook here.
The notebook which processed the original dataset into this one is available here. | [] | [
"TAGS\n#task_categories-text-classification #size_categories-1K<n<10K #language-English #license-mit #region-us \n"
] | [
38
] | [
"passage: TAGS\n#task_categories-text-classification #size_categories-1K<n<10K #language-English #license-mit #region-us \n"
] |
61ea50f52c55527d5b130af76e4af3ccc91deb07 | <p align="center"><img src="https://i.ibb.co/3z86DFV/cygnuscat.png"/><font size="6"> <b>Adjusted to Alpaca from <a href="https://huggingface.co/datasets/mlabonne/chatml_dpo_pairs">mlabonne/chatml_dpo_pairs</a></b> </font></p>
| athirdpath/alpaca_dpo_pairs | [
"license:apache-2.0",
"region:us"
] | 2023-11-30T21:32:24+00:00 | {"license": "apache-2.0"} | 2023-12-01T03:18:22+00:00 | [] | [] | TAGS
#license-apache-2.0 #region-us
| <p align="center"><img src="https://i.URL size="6"> <b>Adjusted to Alpaca from <a href="URL </font></p>
| [] | [
"TAGS\n#license-apache-2.0 #region-us \n"
] | [
14
] | [
"passage: TAGS\n#license-apache-2.0 #region-us \n"
] |
71cebc1756664652935fd5ee1b27bdd0a74074a0 | # Dataset Card for "labelled-deloitte-cloud"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | evkes/labelled-deloitte-cloud | [
"region:us"
] | 2023-11-30T21:47:36+00:00 | {"dataset_info": {"features": [{"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1121994, "num_examples": 654}], "download_size": 411938, "dataset_size": 1121994}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-11-30T21:56:34+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "labelled-deloitte-cloud"
More Information needed | [
"# Dataset Card for \"labelled-deloitte-cloud\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"labelled-deloitte-cloud\"\n\nMore Information needed"
] | [
6,
18
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"labelled-deloitte-cloud\"\n\nMore Information needed"
] |
c52c341593b13cf447e339b37f0fce83643531d8 |
# Make Believe Fake Celebrity News Dataset
| 2nji/makebelieve-480 | [
"license:apache-2.0",
"region:us"
] | 2023-11-30T22:05:30+00:00 | {"license": "apache-2.0", "dataset_info": {"features": [{"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 400355, "num_examples": 480}], "download_size": 245896, "dataset_size": 400355}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-17T09:52:46+00:00 | [] | [] | TAGS
#license-apache-2.0 #region-us
|
# Make Believe Fake Celebrity News Dataset
| [
"# Make Believe Fake Celebrity News Dataset"
] | [
"TAGS\n#license-apache-2.0 #region-us \n",
"# Make Believe Fake Celebrity News Dataset"
] | [
14,
10
] | [
"passage: TAGS\n#license-apache-2.0 #region-us \n# Make Believe Fake Celebrity News Dataset"
] |
e4a8f76e7b040e51bee1e751651b843ff68a70cf | # Dataset Card for "handling_charges_current_v1"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | paul-w-qs/handling_charges_current_v1 | [
"region:us"
] | 2023-11-30T22:28:48+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "JSON_LABEL", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 363782668.918, "num_examples": 1093}], "download_size": 362595207, "dataset_size": 363782668.918}} | 2023-11-30T22:30:49+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "handling_charges_current_v1"
More Information needed | [
"# Dataset Card for \"handling_charges_current_v1\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"handling_charges_current_v1\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"handling_charges_current_v1\"\n\nMore Information needed"
] |
12bdaf34d12cc33f48f8abdce5fcb9585df122ae | # Dataset Card for "midascontrolourbalanced"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | dputilov/midascontrolourbalanced | [
"region:us"
] | 2023-11-30T22:49:30+00:00 | {"dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "conditioning_image", "dtype": "image"}, {"name": "mask", "dtype": "image"}, {"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1537158001.974634, "num_examples": 2725}], "download_size": 1526594996, "dataset_size": 1537158001.974634}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-11-30T22:54:28+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "midascontrolourbalanced"
More Information needed | [
"# Dataset Card for \"midascontrolourbalanced\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"midascontrolourbalanced\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"midascontrolourbalanced\"\n\nMore Information needed"
] |
97b02f709c8b1acef036931a699f3f719ea55a02 | Notes on use:
Portuguese and English Translations of readme are available here.
Partially cleaned and reorganized. Minimal secondhand verification after generation through Google Bard on November 28th 2023. Mistakes are minimal but present, such as tagging of words in supplemental information sometimes using the whole word (ie Noun) and sometimes only a letter or abreviation (ie N) for the same part of speech.
Reccomended for finetuning of smaller models only, such as 12, 7, or 3 B models to create a basic basis for democratizing access to portuguese langauge users, or for inclusion into a much larger training data collections when used for training or finetuning larger models.
The dataset consists of over 2,000 Portuguese-English translation pairs consisting of either a translation pair of words, sentences, or words with definitions, along with (usually) autogenerated G3 Annotations, B- Tags , SRL Annotations, Dependency Parsing Annotations , POS Tagging Annotations, generated using Google Bard on November 28th 2023.
----
2087 observations
*English Readme:*
**Bridging Language Barriers and Empowering Marginalized Communities through Multilingual Parallel Corpora: The English-Portuguese MPC**
In the realm of natural language processing (NLP), the availability of high-quality language data is crucial for developing effective NLP models. However, marginalized indigenous communities, particularly those in Brazil and the Amazon Rainforest, often lack access to NLP tools and resources, hindering their ability to fully engage in the digital age.
To address this disparity, we propose the creation of an English-Portuguese Multilingual Parallel Corpus (MPC), a collection of carefully curated parallel text pairs in English and Portuguese. This resource is specifically designed to democratize access to NLP and promote knowledge exchange between these two languages, fostering cross-cultural communication and empowering marginalized communities.
**Democratizing Access and Fostering Cross-Pollination**
This English-Portuguese MPC, available uncleaned at Solshine/Portuguese-English-Translation-and-NLP-trainingdata-UNCLEANED , serves as a bridge between these two languages, enabling the development of more robust machine translation systems and enriching the availability of Portuguese language resources. For marginalized indigenous communities, this resource holds the potential to:
* **Preserve and revitalize indigenous languages:** By providing a benchmark for Portuguese-English translation, the MPC can facilitate the development of tools that can translate indigenous languages into Portuguese, aiding in language preservation and revitalization efforts.
* **Empower indigenous communities in the digital sphere:** Access to NLP tools trained on the MPC can enable indigenous communities to engage with online resources, participate in social media, and access information in their native language or Portuguese.
* **Promote cross-cultural understanding:** By fostering better communication between indigenous communities and the wider world, the MPC can help bridge cultural divides and promote mutual understanding.
**Addressing the Needs of Marginalized Indigenous Communities**
The English-Portuguese MPC is particularly relevant to marginalized indigenous communities in Brazil and the Amazon Rainforest, where Portuguese serves as the lingua franca and numerous indigenous languages are spoken. By providing a comprehensive resource for Portuguese-English translation, the MPC can empower these communities to:
* **Participate in education and research:** Indigenous students and researchers can access academic materials and engage in scientific research using NLP tools trained on the MPC.
* **Promote cultural heritage and storytelling:** Indigenous communities can utilize the MPC to translate their cultural narratives, folklore, and traditional knowledge into Portuguese, preserving and sharing their heritage with a wider audience.
* **Advocate for their rights and interests:** Indigenous communities can use the MPC to communicate effectively with government officials, NGOs, and the international community, advocating for their rights and interests.
**Sourcing the MPC through Conversation with Bard**
The English-Portuguese MPC was sourced through a combination of conversation with Bard and other sources. Bard's ability to understand and generate human language made it an invaluable tool for generating and refining the dataset, ensuring that the translations were accurate and natural-sounding.
Once the data was collected, it was carefully preprocessed and aligned to ensure that the sentences in English and Portuguese were truly parallel. This was a critical step, as it ensured that the MPC could be used to train NLP models that could accurately translate between the two languages.
**Conclusion**
The English-Portuguese MPC here is a valuable resource for NLP researchers and developers. It can be used to train machine translation systems, develop text summarization tools, and create sentiment analysis models. The MPC can also be used to study the relationship between English and Portuguese, and to develop new NLP algorithms. Moreover, the MPC empowers marginalized communities to preserve their languages, engage in the digital sphere, and advocate for their rights and interests.
----
*Portuguese Readme:*
**Superando barreiras linguísticas e empoderando comunidades marginalizadas por meio de corpora paralelos multilíngues: O MPC inglês-português**
No campo do processamento de linguagem natural (PLN), a disponibilidade de dados linguísticos de alta qualidade é crucial para o desenvolvimento de modelos de PLN eficazes. No entanto, comunidades indígenas marginalizadas, particularmente aquelas no Brasil e na Amazônia, muitas vezes não têm acesso a ferramentas e recursos de PLN, dificultando sua capacidade de se envolver plenamente na era digital.
Para lidar com essa disparidade, propomos a criação de um Corpus Paralelo Multilíngue (MPC) inglês-português, uma coleção de pares de texto paralelos cuidadosamente selecionados em inglês e português. Este recurso é projetado especificamente para democratizar o acesso ao PLN e promover o intercâmbio de conhecimento entre esses dois idiomas, fomentando a comunicação intercultural e empoderando comunidades marginalizadas.
**Democratizando o acesso e fomentando a polinização cruzada**
O MPC inglês-português, Solshine/Portuguese-English-Translation-and-NLP-trainingdata-UNCLEANED , serve como uma ponte entre esses dois idiomas, possibilitando o desenvolvimento de sistemas de tradução automática mais robustos e enriquecendo a disponibilidade de recursos em português. Para comunidades indígenas marginalizadas, este recurso tem o potencial de:
* **Preservar e revitalizar línguas indígenas:** Ao fornecer um benchmark para a tradução português-inglês, o MPC pode facilitar o desenvolvimento de ferramentas que podem traduzir línguas indígenas para o português, auxiliando nos esforços de preservação e revitalização da língua.
* **Empoderar comunidades indígenas na esfera digital:** O acesso a ferramentas de PLN treinadas no MPC pode capacitar comunidades indígenas a se envolver com recursos online, participar de mídias sociais e acessar informações em sua língua nativa ou português.
* **Promover a compreensão intercultural:** Ao promover uma melhor comunicação entre as comunidades indígenas e o mundo em geral, o MPC pode ajudar a superar divisões culturais e promover a compreensão mútua.
**Abordando as necessidades de comunidades indígenas marginalizadas**
O MPC inglês-português é particularmente relevante para comunidades indígenas marginalizadas no Brasil e na Amazônia, onde o português serve como língua franca e inúmeras línguas indígenas são faladas. Ao fornecer um recurso abrangente para tradução português-inglês, o MPC pode capacitar essas comunidades a:
* **Participar de educação e pesquisa:** Estudantes e pesquisadores indígenas podem acessar materiais acadêmicos e se envolver em pesquisas científicas usando ferramentas de PLN treinadas no MPC.
* **Promover o patrimônio cultural e a narrativa:** Comunidades indígenas podem utilizar o MPC para traduzir suas narrativas culturais, folclore e conhecimento tradicional para o português, preservando e compartilhando seu patrimônio com um público mais amplo.
* **Defender seus direitos e interesses:** As comunidades indígenas podem usar o MPC para se comunicar efetivamente com autoridades governamentais, ONGs e a comunidade internacional, defendendo seus direitos e interesses.
**Obtendo o MPC por meio de conversas com Bard**
O MPC inglês-português foi obtido por meio de uma combinação de conversas com Bard e outras fontes. A capacidade do Bard de entender e gerar linguagem humana o tornou uma ferramenta inestimável para gerar e refinar o conjunto de dados, garantindo que as traduções fossem precisas e naturais.
Uma vez que os dados foram coletados, eles foram cuidadosamente pré-processados e alinhados para garantir que as frases em inglês e português fossem verdadeiramente paralelas. Esta foi uma etapa crítica, pois garantiu que o MPC pudesse ser usado para treinar modelos de PLN que pudessem traduzir com precisão entre os dois idiomas.
**Conclusão**
O MPC inglês-português é um recurso valioso para pesquisadores e desenvolvedores de PLN. Ele pode ser usado para treinar sistemas de tradução automática, desenvolver ferramentas de resumo de texto e criar modelos de análise de sentimento. O MPC também pode ser usado para estudar a relação entre inglês e português e desenvolver novos algoritmos de PLN. Além disso, o MPC capacita comunidades marginalizadas a preservar seus idiomas, se envolver na esfera digital e defender seus direitos e interesses.
----
**Notes about the data (currently being updated):**
Inspired by the (November 2023) developments of Ocra2 (albiet much less sophsticated) and the ability for large LLMs to now produce training datasets for smaller (ie 7B or 3B) models to efficienctly learn and distill the fundamentals of the knowledge into themselves.
This is a great Portuguese language dataset, connecting Portuguese into the most widely used and trained language (English), thus democratizing access. You are encouraged to use this in your training to enrich the model's Portuguese.
Most of the table include: G3 Annotations, B- Tags , SRL Annotations, Dependency Parsing Annotations , POS Tagging Annotations
Please note for a large portion of the data: The NER annotations (G3) indicate general entities, while the B- tags indicate specific types of entities (e.g., B-Location, B-Time). The SRL annotations indicate the semantic roles of the constituents in the sentence (e.g., B-Theme, B-Agent, B-Patient, B-Goal). The dependency parsing annotations indicate the grammatical relationships between the words in the sentence. The POS tagging annotations indicate the part-of-speech (e.g., N for noun, V for verb, A for adjective) of each word in the sentence.
The main source of the data is generated through structured questions to Google Bard in the final week of November 2023, and many of these generating queries can be found as title names of individual small tables. Much of that portion was structured explicitly by having the prompt instructions including the previous paragraph's data structure explanation. This resulted in potentially much more useful data about the sentences or words from an NLP perspective, albiet with more inconsistency and minor errors, and even information entered occasionally into the wrong field, in those NLP related fields mentioned above.
This data is mostly uncleaned and should be used with the understanding that it was largely uncleaned and gathered from various sources. The data in the tables in this dataset has inherently been filtered by the guardrails present in Bard and through careful observation of the data (rejecting obviously errorous generations) as it was being generated by Bard and uploaded to, minimally processed the dataset. This is a disclaimer of any possible error or omission, and the dataset should be useful with this understanding.
Some of the tables or observations are missing entries for some of those fields (with the exception of a Portuguese term or sentence, which is present in every observation) especially POS Tagging Annotations, and Dependency Parsing Annotations, usually due to reaching Bard's data limit in it's public browser version (as of November 28th 2023.)
Strong focus towards moral compass and ethical real-world problems solving, as well as Indigenous Knowledge Systems, Climate Change, Science, STEM, intellectual property from a Copyleft perspective, some simple aspects of law, Indigenous Archeology, Educational Philosophy, and basic Vocabulary.
| Solshine/Portuguese-English-Vocab-PartiallyTransformed | [
"license:mit",
"region:us"
] | 2023-11-30T23:07:03+00:00 | {"license": "mit"} | 2023-12-02T19:18:22+00:00 | [] | [] | TAGS
#license-mit #region-us
| Notes on use:
Portuguese and English Translations of readme are available here.
Partially cleaned and reorganized. Minimal secondhand verification after generation through Google Bard on November 28th 2023. Mistakes are minimal but present, such as tagging of words in supplemental information sometimes using the whole word (ie Noun) and sometimes only a letter or abreviation (ie N) for the same part of speech.
Reccomended for finetuning of smaller models only, such as 12, 7, or 3 B models to create a basic basis for democratizing access to portuguese langauge users, or for inclusion into a much larger training data collections when used for training or finetuning larger models.
The dataset consists of over 2,000 Portuguese-English translation pairs consisting of either a translation pair of words, sentences, or words with definitions, along with (usually) autogenerated G3 Annotations, B- Tags , SRL Annotations, Dependency Parsing Annotations , POS Tagging Annotations, generated using Google Bard on November 28th 2023.
----
2087 observations
*English Readme:*
Bridging Language Barriers and Empowering Marginalized Communities through Multilingual Parallel Corpora: The English-Portuguese MPC
In the realm of natural language processing (NLP), the availability of high-quality language data is crucial for developing effective NLP models. However, marginalized indigenous communities, particularly those in Brazil and the Amazon Rainforest, often lack access to NLP tools and resources, hindering their ability to fully engage in the digital age.
To address this disparity, we propose the creation of an English-Portuguese Multilingual Parallel Corpus (MPC), a collection of carefully curated parallel text pairs in English and Portuguese. This resource is specifically designed to democratize access to NLP and promote knowledge exchange between these two languages, fostering cross-cultural communication and empowering marginalized communities.
Democratizing Access and Fostering Cross-Pollination
This English-Portuguese MPC, available uncleaned at Solshine/Portuguese-English-Translation-and-NLP-trainingdata-UNCLEANED , serves as a bridge between these two languages, enabling the development of more robust machine translation systems and enriching the availability of Portuguese language resources. For marginalized indigenous communities, this resource holds the potential to:
* Preserve and revitalize indigenous languages: By providing a benchmark for Portuguese-English translation, the MPC can facilitate the development of tools that can translate indigenous languages into Portuguese, aiding in language preservation and revitalization efforts.
* Empower indigenous communities in the digital sphere: Access to NLP tools trained on the MPC can enable indigenous communities to engage with online resources, participate in social media, and access information in their native language or Portuguese.
* Promote cross-cultural understanding: By fostering better communication between indigenous communities and the wider world, the MPC can help bridge cultural divides and promote mutual understanding.
Addressing the Needs of Marginalized Indigenous Communities
The English-Portuguese MPC is particularly relevant to marginalized indigenous communities in Brazil and the Amazon Rainforest, where Portuguese serves as the lingua franca and numerous indigenous languages are spoken. By providing a comprehensive resource for Portuguese-English translation, the MPC can empower these communities to:
* Participate in education and research: Indigenous students and researchers can access academic materials and engage in scientific research using NLP tools trained on the MPC.
* Promote cultural heritage and storytelling: Indigenous communities can utilize the MPC to translate their cultural narratives, folklore, and traditional knowledge into Portuguese, preserving and sharing their heritage with a wider audience.
* Advocate for their rights and interests: Indigenous communities can use the MPC to communicate effectively with government officials, NGOs, and the international community, advocating for their rights and interests.
Sourcing the MPC through Conversation with Bard
The English-Portuguese MPC was sourced through a combination of conversation with Bard and other sources. Bard's ability to understand and generate human language made it an invaluable tool for generating and refining the dataset, ensuring that the translations were accurate and natural-sounding.
Once the data was collected, it was carefully preprocessed and aligned to ensure that the sentences in English and Portuguese were truly parallel. This was a critical step, as it ensured that the MPC could be used to train NLP models that could accurately translate between the two languages.
Conclusion
The English-Portuguese MPC here is a valuable resource for NLP researchers and developers. It can be used to train machine translation systems, develop text summarization tools, and create sentiment analysis models. The MPC can also be used to study the relationship between English and Portuguese, and to develop new NLP algorithms. Moreover, the MPC empowers marginalized communities to preserve their languages, engage in the digital sphere, and advocate for their rights and interests.
----
*Portuguese Readme:*
Superando barreiras linguísticas e empoderando comunidades marginalizadas por meio de corpora paralelos multilíngues: O MPC inglês-português
No campo do processamento de linguagem natural (PLN), a disponibilidade de dados linguísticos de alta qualidade é crucial para o desenvolvimento de modelos de PLN eficazes. No entanto, comunidades indígenas marginalizadas, particularmente aquelas no Brasil e na Amazônia, muitas vezes não têm acesso a ferramentas e recursos de PLN, dificultando sua capacidade de se envolver plenamente na era digital.
Para lidar com essa disparidade, propomos a criação de um Corpus Paralelo Multilíngue (MPC) inglês-português, uma coleção de pares de texto paralelos cuidadosamente selecionados em inglês e português. Este recurso é projetado especificamente para democratizar o acesso ao PLN e promover o intercâmbio de conhecimento entre esses dois idiomas, fomentando a comunicação intercultural e empoderando comunidades marginalizadas.
Democratizando o acesso e fomentando a polinização cruzada
O MPC inglês-português, Solshine/Portuguese-English-Translation-and-NLP-trainingdata-UNCLEANED , serve como uma ponte entre esses dois idiomas, possibilitando o desenvolvimento de sistemas de tradução automática mais robustos e enriquecendo a disponibilidade de recursos em português. Para comunidades indígenas marginalizadas, este recurso tem o potencial de:
* Preservar e revitalizar línguas indígenas: Ao fornecer um benchmark para a tradução português-inglês, o MPC pode facilitar o desenvolvimento de ferramentas que podem traduzir línguas indígenas para o português, auxiliando nos esforços de preservação e revitalização da língua.
* Empoderar comunidades indígenas na esfera digital: O acesso a ferramentas de PLN treinadas no MPC pode capacitar comunidades indígenas a se envolver com recursos online, participar de mídias sociais e acessar informações em sua língua nativa ou português.
* Promover a compreensão intercultural: Ao promover uma melhor comunicação entre as comunidades indígenas e o mundo em geral, o MPC pode ajudar a superar divisões culturais e promover a compreensão mútua.
Abordando as necessidades de comunidades indígenas marginalizadas
O MPC inglês-português é particularmente relevante para comunidades indígenas marginalizadas no Brasil e na Amazônia, onde o português serve como língua franca e inúmeras línguas indígenas são faladas. Ao fornecer um recurso abrangente para tradução português-inglês, o MPC pode capacitar essas comunidades a:
* Participar de educação e pesquisa: Estudantes e pesquisadores indígenas podem acessar materiais acadêmicos e se envolver em pesquisas científicas usando ferramentas de PLN treinadas no MPC.
* Promover o patrimônio cultural e a narrativa: Comunidades indígenas podem utilizar o MPC para traduzir suas narrativas culturais, folclore e conhecimento tradicional para o português, preservando e compartilhando seu patrimônio com um público mais amplo.
* Defender seus direitos e interesses: As comunidades indígenas podem usar o MPC para se comunicar efetivamente com autoridades governamentais, ONGs e a comunidade internacional, defendendo seus direitos e interesses.
Obtendo o MPC por meio de conversas com Bard
O MPC inglês-português foi obtido por meio de uma combinação de conversas com Bard e outras fontes. A capacidade do Bard de entender e gerar linguagem humana o tornou uma ferramenta inestimável para gerar e refinar o conjunto de dados, garantindo que as traduções fossem precisas e naturais.
Uma vez que os dados foram coletados, eles foram cuidadosamente pré-processados e alinhados para garantir que as frases em inglês e português fossem verdadeiramente paralelas. Esta foi uma etapa crítica, pois garantiu que o MPC pudesse ser usado para treinar modelos de PLN que pudessem traduzir com precisão entre os dois idiomas.
Conclusão
O MPC inglês-português é um recurso valioso para pesquisadores e desenvolvedores de PLN. Ele pode ser usado para treinar sistemas de tradução automática, desenvolver ferramentas de resumo de texto e criar modelos de análise de sentimento. O MPC também pode ser usado para estudar a relação entre inglês e português e desenvolver novos algoritmos de PLN. Além disso, o MPC capacita comunidades marginalizadas a preservar seus idiomas, se envolver na esfera digital e defender seus direitos e interesses.
----
Notes about the data (currently being updated):
Inspired by the (November 2023) developments of Ocra2 (albiet much less sophsticated) and the ability for large LLMs to now produce training datasets for smaller (ie 7B or 3B) models to efficienctly learn and distill the fundamentals of the knowledge into themselves.
This is a great Portuguese language dataset, connecting Portuguese into the most widely used and trained language (English), thus democratizing access. You are encouraged to use this in your training to enrich the model's Portuguese.
Most of the table include: G3 Annotations, B- Tags , SRL Annotations, Dependency Parsing Annotations , POS Tagging Annotations
Please note for a large portion of the data: The NER annotations (G3) indicate general entities, while the B- tags indicate specific types of entities (e.g., B-Location, B-Time). The SRL annotations indicate the semantic roles of the constituents in the sentence (e.g., B-Theme, B-Agent, B-Patient, B-Goal). The dependency parsing annotations indicate the grammatical relationships between the words in the sentence. The POS tagging annotations indicate the part-of-speech (e.g., N for noun, V for verb, A for adjective) of each word in the sentence.
The main source of the data is generated through structured questions to Google Bard in the final week of November 2023, and many of these generating queries can be found as title names of individual small tables. Much of that portion was structured explicitly by having the prompt instructions including the previous paragraph's data structure explanation. This resulted in potentially much more useful data about the sentences or words from an NLP perspective, albiet with more inconsistency and minor errors, and even information entered occasionally into the wrong field, in those NLP related fields mentioned above.
This data is mostly uncleaned and should be used with the understanding that it was largely uncleaned and gathered from various sources. The data in the tables in this dataset has inherently been filtered by the guardrails present in Bard and through careful observation of the data (rejecting obviously errorous generations) as it was being generated by Bard and uploaded to, minimally processed the dataset. This is a disclaimer of any possible error or omission, and the dataset should be useful with this understanding.
Some of the tables or observations are missing entries for some of those fields (with the exception of a Portuguese term or sentence, which is present in every observation) especially POS Tagging Annotations, and Dependency Parsing Annotations, usually due to reaching Bard's data limit in it's public browser version (as of November 28th 2023.)
Strong focus towards moral compass and ethical real-world problems solving, as well as Indigenous Knowledge Systems, Climate Change, Science, STEM, intellectual property from a Copyleft perspective, some simple aspects of law, Indigenous Archeology, Educational Philosophy, and basic Vocabulary.
| [] | [
"TAGS\n#license-mit #region-us \n"
] | [
11
] | [
"passage: TAGS\n#license-mit #region-us \n"
] |
12f0baa6526d2e52ba7844035b35ff4891deae34 |
---
About Dataset
---
This dataset was scrapped from the MedQuAD repository and then converted to a csv file. This file contains multiple categories such as:
1. Cancer
2. Senior Helth
3. Growth Hormones and Receptors
4. Heart, Lungs and Blood
5. Genetic and Rare Diseases
6. Disease Control and Prevention
7. Neurological Disorders and Stroke
8. Diabetes and Rare Diseases
9. Other
The above dataset belongs to the https://github.com/abachaa/MedQuAD.git repository, so please refer to them for more information. I have attached their README file information for ease of data literacy.
This dataset was splitted 80% for the train and 20% for the test for each topic separetely and then merged to one file.
---
MedQuAD: Medical Question Answering Dataset
---
MedQuAD includes 47,457 medical question-answer pairs created from 12 NIH websites (e.g. cancer.gov, niddk.nih.gov, GARD, MedlinePlus Health Topics). The collection covers 37 question types (e.g. Treatment, Diagnosis, Side Effects) associated with diseases, drugs and other medical entities such as tests.
They included additional annotations in the XML files, that could be used for diverse IR and NLP tasks, such as the question type, the question focus, its syonyms, its UMLS Concept Unique Identifier (CUI) and Semantic Type.
They added the category of the question focus (Disease, Drug or Other) in the 4 MedlinePlus collections. All other collections are about diseases.
The paper cited below describes the collection, the construction method as well as its use and evaluation within a medical question answering system.
N.B. They removed the answers from 3 subsets to respect the MedlinePlus copyright (https://medlineplus.gov/copyright.html):
(1) A.D.A.M. Medical Encyclopedia, (2) MedlinePlus Drug information, and (3) MedlinePlus Herbal medicine and supplement information.
-- They kept all the other information including the URLs in case you want to crawl the answers. Please contact them if you have any questions.
---
QA Test Collection
---
They used the test questions of the TREC-2017 LiveQA medical task: https://github.com/abachaa/LiveQA_MedicalTask_TREC2017/tree/master/TestDataset.
As described in their BMC paper, they have manually judged the answers retrieved by the IR and QA systems from the MedQuAD collection.
They used the same judgment scores as the LiveQA Track: 1-Incorrect, 2-Related, 3-Incomplete, and 4-Excellent.
-- Format of the qrels file: Question_ID judgment Answer_ID
The QA test collection contains 2,479 judged answers that can be used to evaluate the performance of IR & QA systems on the LiveQA-Med test questions: https://github.com/abachaa/MedQuAD/blob/master/QA-TestSet-LiveQA-Med-Qrels-2479-Answers.zip
---
Reference
---
If you use the MedQuAD dataset and/or the collection of 2,479 judged answers, please cite the following paper: "A Question-Entailment Approach to Question Answering". Asma Ben Abacha and Dina Demner-Fushman. BMC Bioinformatics, 2019.
@ARTICLE{BenAbacha-BMC-2019,
author = {Asma {Ben Abacha} and Dina Demner{-}Fushman},
title = {A Question-Entailment Approach to Question Answering},
journal = {{BMC} Bioinform.},
volume = {20},
number = {1},
pages = {511:1--511:23},
year = {2019},
url = {https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3119-4}
}
---
License
---
The MedQuAD dataset is published under a Creative Commons Attribution 4.0 International Licence (CC BY). https://creativecommons.org/licenses/by/4.0/ | GonzaloValdenebro/MedicalQuestionAnsweringDataset | [
"license:mit",
"region:us"
] | 2023-11-30T23:32:08+00:00 | {"license": "mit"} | 2023-12-05T04:06:41+00:00 | [] | [] | TAGS
#license-mit #region-us
|
---
About Dataset
---
This dataset was scrapped from the MedQuAD repository and then converted to a csv file. This file contains multiple categories such as:
1. Cancer
2. Senior Helth
3. Growth Hormones and Receptors
4. Heart, Lungs and Blood
5. Genetic and Rare Diseases
6. Disease Control and Prevention
7. Neurological Disorders and Stroke
8. Diabetes and Rare Diseases
9. Other
The above dataset belongs to the URL repository, so please refer to them for more information. I have attached their README file information for ease of data literacy.
This dataset was splitted 80% for the train and 20% for the test for each topic separetely and then merged to one file.
---
MedQuAD: Medical Question Answering Dataset
---
MedQuAD includes 47,457 medical question-answer pairs created from 12 NIH websites (e.g. URL, URL, GARD, MedlinePlus Health Topics). The collection covers 37 question types (e.g. Treatment, Diagnosis, Side Effects) associated with diseases, drugs and other medical entities such as tests.
They included additional annotations in the XML files, that could be used for diverse IR and NLP tasks, such as the question type, the question focus, its syonyms, its UMLS Concept Unique Identifier (CUI) and Semantic Type.
They added the category of the question focus (Disease, Drug or Other) in the 4 MedlinePlus collections. All other collections are about diseases.
The paper cited below describes the collection, the construction method as well as its use and evaluation within a medical question answering system.
N.B. They removed the answers from 3 subsets to respect the MedlinePlus copyright (URL
(1) A.D.A.M. Medical Encyclopedia, (2) MedlinePlus Drug information, and (3) MedlinePlus Herbal medicine and supplement information.
-- They kept all the other information including the URLs in case you want to crawl the answers. Please contact them if you have any questions.
---
QA Test Collection
---
They used the test questions of the TREC-2017 LiveQA medical task: URL
As described in their BMC paper, they have manually judged the answers retrieved by the IR and QA systems from the MedQuAD collection.
They used the same judgment scores as the LiveQA Track: 1-Incorrect, 2-Related, 3-Incomplete, and 4-Excellent.
-- Format of the qrels file: Question_ID judgment Answer_ID
The QA test collection contains 2,479 judged answers that can be used to evaluate the performance of IR & QA systems on the LiveQA-Med test questions: URL
---
Reference
---
If you use the MedQuAD dataset and/or the collection of 2,479 judged answers, please cite the following paper: "A Question-Entailment Approach to Question Answering". Asma Ben Abacha and Dina Demner-Fushman. BMC Bioinformatics, 2019.
@ARTICLE{BenAbacha-BMC-2019,
author = {Asma {Ben Abacha} and Dina Demner{-}Fushman},
title = {A Question-Entailment Approach to Question Answering},
journal = {{BMC} Bioinform.},
volume = {20},
number = {1},
pages = {511:1--511:23},
year = {2019},
url = {URL
}
---
License
---
The MedQuAD dataset is published under a Creative Commons Attribution 4.0 International Licence (CC BY). URL | [] | [
"TAGS\n#license-mit #region-us \n"
] | [
11
] | [
"passage: TAGS\n#license-mit #region-us \n"
] |
8efa0bb5b3e5e41d2b15f85c354d5c52f7997f0c | # Dataset Card for "conandoyle_cue_scope"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | joey234/conandoyle_cue_scope | [
"region:us"
] | 2023-11-30T23:37:42+00:00 | {"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "cue", "sequence": "int64"}, {"name": "scope", "sequence": "int64"}], "splits": [{"name": "train", "num_bytes": 105721, "num_examples": 235}], "download_size": 21262, "dataset_size": 105721}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-11-30T23:37:44+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "conandoyle_cue_scope"
More Information needed | [
"# Dataset Card for \"conandoyle_cue_scope\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"conandoyle_cue_scope\"\n\nMore Information needed"
] | [
6,
18
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"conandoyle_cue_scope\"\n\nMore Information needed"
] |
b3b8ca200e5d77ba99a8aa6d301e038dfd21ed49 | # Dataset Card for "context_extension-mistral-16k"
!!! idx=27710 has length 32717
remove it with dataset['train'] = dataset['train'].select((i for i in range(len(dataset['train'])) if i != 27710))
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | sade-adrien/context_extension-mistral-16k | [
"region:us"
] | 2023-11-30T23:50:28+00:00 | {"dataset_info": {"features": [{"name": "raw_content", "dtype": "string"}, {"name": "input_ids", "sequence": "int32"}, {"name": "attention_mask", "sequence": "int8"}, {"name": "labels", "sequence": "int64"}], "splits": [{"name": "train", "num_bytes": 13206603486, "num_examples": 30000}], "download_size": 5395605016, "dataset_size": 13206603486}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T18:16:36+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "context_extension-mistral-16k"
!!! idx=27710 has length 32717
remove it with dataset['train'] = dataset['train'].select((i for i in range(len(dataset['train'])) if i != 27710))
More Information needed | [
"# Dataset Card for \"context_extension-mistral-16k\"\n\n!!! idx=27710 has length 32717\n\nremove it with dataset['train'] = dataset['train'].select((i for i in range(len(dataset['train'])) if i != 27710))\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"context_extension-mistral-16k\"\n\n!!! idx=27710 has length 32717\n\nremove it with dataset['train'] = dataset['train'].select((i for i in range(len(dataset['train'])) if i != 27710))\n\nMore Information needed"
] | [
6,
82
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"context_extension-mistral-16k\"\n\n!!! idx=27710 has length 32717\n\nremove it with dataset['train'] = dataset['train'].select((i for i in range(len(dataset['train'])) if i != 27710))\n\nMore Information needed"
] |
c777517ee1ad675f0fea441354f8a756c002c7b8 |
# Description of the Dataset
This release integrates the entire data sequence utilized in the CrystalCoder training. It encompasses data sequences from the three pre-training stages, combining information from two prior works: the [SlimPajama dataset](https://huggingface.co/datasets/cerebras/SlimPajama-627B) and [StarCoder](https://huggingface.co/datasets/bigcode/starcoderdata), totaling approximately 1300 billion tokens. These tokens are distributed across three stages, each with distinct weights.
## Stage 1
During this initial stage, half of the [SlimPajama data](https://huggingface.co/datasets/cerebras/SlimPajama-627B) is utilized, equivalent to approximately 345 billion tokens.
## Stage 2
In the second stage, the remaining half of the [SlimPajama data](https://huggingface.co/datasets/cerebras/SlimPajama-627B) is employed, along with two epochs of [StarCoder data](https://huggingface.co/datasets/bigcode/starcoderdata). For the StarCoder data, we apply [FIM augmentation](https://arxiv.org/abs/2207.14255) with an FIM rate of 0.9 and an SPM rate of 0.5. The total token count for this stage is calculated as 0.5 * 690 + 2 * 291, resulting in 927 billion tokens.
## Stage 3
The third stage involves reusing Python and web-related data from the [StarCoder data](https://huggingface.co/datasets/bigcode/starcoderdata), including HTML, CSS, and JavaScript. This data is utilized for training over three epochs, with the application of FIM at a rate of 0.3 alongside an SPM rate of 0.5. The total token count for this stage is 100 billion. Additionally, a small portion of the SlimPajama dataset, excluding the Github part, is also reused, contributing around 10 billion tokens.
### Instruction tuning (Stage 3a)
To enhance the model's proficiency in real chat scenarios, we utilize a diverse set of instruction tuning datasets, totaling approximately 1 billion tokens. Specifically, our data include [OASST1-guanaco](https://huggingface.co/datasets/openaccess-ai-collective/oasst1-guanaco-extended-sharegpt), [SlimOrca](https://huggingface.co/datasets/Open-Orca/SlimOrca), [ShareGPT_V4.3](https://huggingface.co/datasets/Aeala/ShareGPT_Vicuna_unfiltered), [Evol-ShareGPT](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k), [CodeAlpaca](https://huggingface.co/datasets/lucasmccabe-lmi/CodeAlpaca-20k), [Rosetta Code](https://github.com/sahil280114/codealpaca/blob/master/data/rosetta_alpaca.json), [Evol-CodeAlpaca 1](https://huggingface.co/datasets/theblackcat102/evol-codealpaca-v1), [Evol-CodeAlpaca 2](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1), and a self-generated dataset centered on website creation through the [Alpaca](https://github.com/tatsu-lab/stanford_alpaca) pipeline. We will release the full dataset soon.
The detailed breakdown of the tokens is as followed:
# Primary Usage
This dataset serves as the foundation for training CrystalCoder and supports further reproduction. For training from scratch, please refer to our [training codes](https://github.com/LLM360/crystalcoder-train). For training from middle checkpoints, please load the dataloader states in checkpoints and follow [this tutorial](https://docs.cerebras.net/en/latest/wsc/tutorials/dataloader-checkpointing.html).
# License
Pretraining data in langauge model mostly comes from a collection of data sources with various licenses. Any use of all or part of the data here must abide by the terms of the original licenses, including attribution clauses when relevant. We refer users to [SlimPajama dataset](https://huggingface.co/datasets/cerebras/SlimPajama-627B) and [StarCoder](https://huggingface.co/datasets/bigcode/starcoderdata) for detailed license attribution.
| LLM360/CrystalCoderDatasets | [
"language:en",
"pretrained",
"arxiv:2207.14255",
"region:us"
] | 2023-11-30T23:57:32+00:00 | {"language": ["en"], "tags": ["pretrained"]} | 2024-02-06T16:24:04+00:00 | [
"2207.14255"
] | [
"en"
] | TAGS
#language-English #pretrained #arxiv-2207.14255 #region-us
|
# Description of the Dataset
This release integrates the entire data sequence utilized in the CrystalCoder training. It encompasses data sequences from the three pre-training stages, combining information from two prior works: the SlimPajama dataset and StarCoder, totaling approximately 1300 billion tokens. These tokens are distributed across three stages, each with distinct weights.
## Stage 1
During this initial stage, half of the SlimPajama data is utilized, equivalent to approximately 345 billion tokens.
## Stage 2
In the second stage, the remaining half of the SlimPajama data is employed, along with two epochs of StarCoder data. For the StarCoder data, we apply FIM augmentation with an FIM rate of 0.9 and an SPM rate of 0.5. The total token count for this stage is calculated as 0.5 * 690 + 2 * 291, resulting in 927 billion tokens.
## Stage 3
The third stage involves reusing Python and web-related data from the StarCoder data, including HTML, CSS, and JavaScript. This data is utilized for training over three epochs, with the application of FIM at a rate of 0.3 alongside an SPM rate of 0.5. The total token count for this stage is 100 billion. Additionally, a small portion of the SlimPajama dataset, excluding the Github part, is also reused, contributing around 10 billion tokens.
### Instruction tuning (Stage 3a)
To enhance the model's proficiency in real chat scenarios, we utilize a diverse set of instruction tuning datasets, totaling approximately 1 billion tokens. Specifically, our data include OASST1-guanaco, SlimOrca, ShareGPT_V4.3, Evol-ShareGPT, CodeAlpaca, Rosetta Code, Evol-CodeAlpaca 1, Evol-CodeAlpaca 2, and a self-generated dataset centered on website creation through the Alpaca pipeline. We will release the full dataset soon.
The detailed breakdown of the tokens is as followed:
!data split
# Primary Usage
This dataset serves as the foundation for training CrystalCoder and supports further reproduction. For training from scratch, please refer to our training codes. For training from middle checkpoints, please load the dataloader states in checkpoints and follow this tutorial.
# License
Pretraining data in langauge model mostly comes from a collection of data sources with various licenses. Any use of all or part of the data here must abide by the terms of the original licenses, including attribution clauses when relevant. We refer users to SlimPajama dataset and StarCoder for detailed license attribution.
| [
"# Description of the Dataset\n\nThis release integrates the entire data sequence utilized in the CrystalCoder training. It encompasses data sequences from the three pre-training stages, combining information from two prior works: the SlimPajama dataset and StarCoder, totaling approximately 1300 billion tokens. These tokens are distributed across three stages, each with distinct weights.",
"## Stage 1\nDuring this initial stage, half of the SlimPajama data is utilized, equivalent to approximately 345 billion tokens.",
"## Stage 2\nIn the second stage, the remaining half of the SlimPajama data is employed, along with two epochs of StarCoder data. For the StarCoder data, we apply FIM augmentation with an FIM rate of 0.9 and an SPM rate of 0.5. The total token count for this stage is calculated as 0.5 * 690 + 2 * 291, resulting in 927 billion tokens.",
"## Stage 3\nThe third stage involves reusing Python and web-related data from the StarCoder data, including HTML, CSS, and JavaScript. This data is utilized for training over three epochs, with the application of FIM at a rate of 0.3 alongside an SPM rate of 0.5. The total token count for this stage is 100 billion. Additionally, a small portion of the SlimPajama dataset, excluding the Github part, is also reused, contributing around 10 billion tokens.",
"### Instruction tuning (Stage 3a)\n\nTo enhance the model's proficiency in real chat scenarios, we utilize a diverse set of instruction tuning datasets, totaling approximately 1 billion tokens. Specifically, our data include OASST1-guanaco, SlimOrca, ShareGPT_V4.3, Evol-ShareGPT, CodeAlpaca, Rosetta Code, Evol-CodeAlpaca 1, Evol-CodeAlpaca 2, and a self-generated dataset centered on website creation through the Alpaca pipeline. We will release the full dataset soon.\n\nThe detailed breakdown of the tokens is as followed:\n\n!data split",
"# Primary Usage\n\nThis dataset serves as the foundation for training CrystalCoder and supports further reproduction. For training from scratch, please refer to our training codes. For training from middle checkpoints, please load the dataloader states in checkpoints and follow this tutorial.",
"# License\nPretraining data in langauge model mostly comes from a collection of data sources with various licenses. Any use of all or part of the data here must abide by the terms of the original licenses, including attribution clauses when relevant. We refer users to SlimPajama dataset and StarCoder for detailed license attribution."
] | [
"TAGS\n#language-English #pretrained #arxiv-2207.14255 #region-us \n",
"# Description of the Dataset\n\nThis release integrates the entire data sequence utilized in the CrystalCoder training. It encompasses data sequences from the three pre-training stages, combining information from two prior works: the SlimPajama dataset and StarCoder, totaling approximately 1300 billion tokens. These tokens are distributed across three stages, each with distinct weights.",
"## Stage 1\nDuring this initial stage, half of the SlimPajama data is utilized, equivalent to approximately 345 billion tokens.",
"## Stage 2\nIn the second stage, the remaining half of the SlimPajama data is employed, along with two epochs of StarCoder data. For the StarCoder data, we apply FIM augmentation with an FIM rate of 0.9 and an SPM rate of 0.5. The total token count for this stage is calculated as 0.5 * 690 + 2 * 291, resulting in 927 billion tokens.",
"## Stage 3\nThe third stage involves reusing Python and web-related data from the StarCoder data, including HTML, CSS, and JavaScript. This data is utilized for training over three epochs, with the application of FIM at a rate of 0.3 alongside an SPM rate of 0.5. The total token count for this stage is 100 billion. Additionally, a small portion of the SlimPajama dataset, excluding the Github part, is also reused, contributing around 10 billion tokens.",
"### Instruction tuning (Stage 3a)\n\nTo enhance the model's proficiency in real chat scenarios, we utilize a diverse set of instruction tuning datasets, totaling approximately 1 billion tokens. Specifically, our data include OASST1-guanaco, SlimOrca, ShareGPT_V4.3, Evol-ShareGPT, CodeAlpaca, Rosetta Code, Evol-CodeAlpaca 1, Evol-CodeAlpaca 2, and a self-generated dataset centered on website creation through the Alpaca pipeline. We will release the full dataset soon.\n\nThe detailed breakdown of the tokens is as followed:\n\n!data split",
"# Primary Usage\n\nThis dataset serves as the foundation for training CrystalCoder and supports further reproduction. For training from scratch, please refer to our training codes. For training from middle checkpoints, please load the dataloader states in checkpoints and follow this tutorial.",
"# License\nPretraining data in langauge model mostly comes from a collection of data sources with various licenses. Any use of all or part of the data here must abide by the terms of the original licenses, including attribution clauses when relevant. We refer users to SlimPajama dataset and StarCoder for detailed license attribution."
] | [
22,
88,
28,
94,
114,
152,
62,
71
] | [
"passage: TAGS\n#language-English #pretrained #arxiv-2207.14255 #region-us \n# Description of the Dataset\n\nThis release integrates the entire data sequence utilized in the CrystalCoder training. It encompasses data sequences from the three pre-training stages, combining information from two prior works: the SlimPajama dataset and StarCoder, totaling approximately 1300 billion tokens. These tokens are distributed across three stages, each with distinct weights.## Stage 1\nDuring this initial stage, half of the SlimPajama data is utilized, equivalent to approximately 345 billion tokens.## Stage 2\nIn the second stage, the remaining half of the SlimPajama data is employed, along with two epochs of StarCoder data. For the StarCoder data, we apply FIM augmentation with an FIM rate of 0.9 and an SPM rate of 0.5. The total token count for this stage is calculated as 0.5 * 690 + 2 * 291, resulting in 927 billion tokens.## Stage 3\nThe third stage involves reusing Python and web-related data from the StarCoder data, including HTML, CSS, and JavaScript. This data is utilized for training over three epochs, with the application of FIM at a rate of 0.3 alongside an SPM rate of 0.5. The total token count for this stage is 100 billion. Additionally, a small portion of the SlimPajama dataset, excluding the Github part, is also reused, contributing around 10 billion tokens.### Instruction tuning (Stage 3a)\n\nTo enhance the model's proficiency in real chat scenarios, we utilize a diverse set of instruction tuning datasets, totaling approximately 1 billion tokens. Specifically, our data include OASST1-guanaco, SlimOrca, ShareGPT_V4.3, Evol-ShareGPT, CodeAlpaca, Rosetta Code, Evol-CodeAlpaca 1, Evol-CodeAlpaca 2, and a self-generated dataset centered on website creation through the Alpaca pipeline. We will release the full dataset soon.\n\nThe detailed breakdown of the tokens is as followed:\n\n!data split"
] |
315e1440595ac79a54db2eefc1c1f3c52f825741 |
### **UniIR: Training and Benchmarking Universal Multimodal Information Retrievers**
[**🌐 Homepage**](https://tiger-ai-lab.github.io/UniIR/) | [**🤗 Paper**](https://huggingface.co/papers/2311.17136) | [**📖 arXiv**](https://arxiv.org/pdf/2311.17136.pdf) | [**GitHub**](https://github.com/TIGER-AI-Lab/UniIR)
## 🔔News
- **🔥[2023-12-21]: Our M-BEIR Benchmark is now available for use.**
## **Dataset Summary**
**M-BEIR**, the **M**ultimodal **BE**nchmark for **I**nstructed **R**etrieval, is a comprehensive large-scale retrieval benchmark designed to train and evaluate unified multimodal retrieval models (**UniIR models**).
The M-BEIR benchmark comprises eight multimodal retrieval tasks and ten datasets from a variety of domains and sources.
Each task is accompanied by human-authored instructions, encompassing 1.5 million queries and a pool of 5.6 million retrieval candidates in total.
## **Dataset Structure Overview**
The M-BEIR dataset is structured into five primary components: Query Data, Candidate Pool, Instructions, Qrels, and Images.
### Query Data
Below is the directory structure for the query data:
```
query/
│
├── train/
│ ├── mbeir_cirr_train.jsonl
│ ├── mbeir_edis_train.jsonl
│ ...
├── union_train/
│ └── mbeir_union_up_train.jsonl
├── val/
│ ├── mbeir_visualnews_task0_val.jsonl
│ ├── mbeir_visualnews_task3_val.jsonl
│ ...
└── test/
├── mbeir_visualnews_task0_test.jsonl
├── mbeir_visualnews_task3_test.jsonl
...
```
`train`: Contains all the training data from 8 different datasets formatted in the M-BEIR style.
`mbeir_union_up_train.jsonl`: This file is the default training data for in-batch contrastive training specifically designed for UniIR models.
It aggregates all the data from the train directory and datasets with relatively smaller sizes have been upsampled to balance the training process.
`val`: Contains separate files for validation queries, organized by task.
`test`: Contains separate files for test queries, organized by task.
Every M-BEIR query instance has at least one positive candidate data and possibly no negative candidate data
Each line in a Query Data file represents a unique query. The structure of each query JSON object is as follows::
```json
{
"qid": "A unique identifier formatted as {dataset_id}:{query_id}",
"query_txt": "The text component of the query",
"query_img_path": "The file path to the associated query image",
"query_modality": "The modality type of the query (text, image or text,image)",
"query_src_content": "Additional content from the original dataset, presented as a string by json.dumps()",
"pos_cand_list": [
{
"did": "A unique identifier formatted as {dataset_id}:{doc_id}"
}
// ... more positive candidates
],
"neg_cand_list": [
{
"did": "A unique identifier formatted as {dataset_id}:{doc_id}"
}
// ... more negative candidates
]
}
```
### Candidate Pool
The Candidate Pool contains potential matching documents for the queries.
#### M-BEIR_5.6M
Within the global directory, the default retrieval setting requires models to retrieve positive candidates from a heterogeneous pool encompassing various modalities and domains.
The M-BEIR's global candidate pool, comprising 5.6 million candidates, includes the retrieval corpus from all tasks and datasets.
#### M-BEIR_local
Within the local directory, we provide dataset-task-specific pool as M-BEIR_local. Dataset-task-specific pool contains homogeneous candidates that originate from by the original dataset.
Below is the directory structure for the candidate pool:
```
cand_pool/
│
├── global/
│ ├── mbeir_union_val_cand_pool.jsonl
│ └──mbeir_union_test_cand_pool.jsonl
│
└── local/
├── mbeir_visualnews_task0_cand_pool.jsonl
├── mbeir_visualnews_task3_cand_pool.jsonl
...
```
The structure of each candidate JSON object in cand_pool file is as follows::
```json
{
"did": "A unique identifier for the document, formatted as {dataset_id}:{doc_id}",
"txt": "The text content of the candidate document",
"img_path": "The file path to the candidate document's image",
"modality": "The modality type of the candidate (e.g., text, image or text,image)",
"src_content": "Additional content from the original dataset, presented as a string by json.dumps()"
}
```
### Instructions
`query_instructions.tsv` contains human-authorized instructions within the UniIR framework. Each task is accompanied by four human-authored instructions. For detailed usage, please refer to [**GitHub Repo**](https://github.com/TIGER-AI-Lab/UniIR).
### Qrels
Within the `qrels` directory, you will find qrels for both the validation and test sets. These files serve the purpose of evaluating UniIR models. For detailed information, please refer to [**GitHub Repo**](https://github.com/TIGER-AI-Lab/UniIR).
## **How to Use**
### Downloading the M-BEIR Dataset
Clone the M-BEIR repo from the current Page.
Ensure that Git LFS (Large File Storage) is installed on your system, as it will download the required data files.
### Decompressing M-BEIR Images
After downloading, you will need to decompress the image files. Follow these steps in your terminal:
```bash
# Navigate to the M-BEIR directory
cd path/to/M-BEIR
# Combine the split tar.gz files into one
sh -c 'cat mbeir_images.tar.gz.part-00 mbeir_images.tar.gz.part-01 mbeir_images.tar.gz.part-02 mbeir_images.tar.gz.part-03 > mbeir_images.tar.gz'
# Extract the images from the tar.gz file
tar -xzf mbeir_images.tar.gz
```
Now, you are ready to use the M-BEIR benchmark.
### Dataloader and Evaluation Pipeline
We offer a dedicated dataloader and evaluation pipeline for the M-BEIR benchmark. Please refer to [**GitHub Repo**](https://github.com/TIGER-AI-Lab/UniIR) for detailed information.
## **Citation**
Please cite our paper if you use our data, model or code.
```
@article{wei2023uniir,
title={UniIR: Training and Benchmarking Universal Multimodal Information Retrievers},
author={Wei, Cong and Chen, Yang and Chen, Haonan and Hu, Hexiang and Zhang, Ge and Fu, Jie and Ritter, Alan and Chen, Wenhu},
journal={arXiv preprint arXiv:2311.17136},
year={2023}
}
```
| TIGER-Lab/M-BEIR | [
"task_categories:text-retrieval",
"task_categories:text-to-image",
"task_categories:image-to-text",
"task_categories:visual-question-answering",
"language:en",
"license:mit",
"arxiv:2311.17136",
"region:us"
] | 2023-12-01T00:08:47+00:00 | {"language": ["en"], "license": "mit", "task_categories": ["text-retrieval", "text-to-image", "image-to-text", "visual-question-answering"], "pretty_name": "M-BEIR", "configs": [{"config_name": "query", "data_files": [{"split": "train", "path": "query/train/*.jsonl"}, {"split": "union_train", "path": "query/union_train/*.jsonl"}, {"split": "val", "path": "query/val/*.jsonl"}, {"split": "test", "path": "query/test/*.jsonl"}]}, {"config_name": "cand_pool", "data_files": [{"split": "mbeir_local", "path": "cand_pool/local/*.jsonl"}, {"split": "mbeir_global", "path": "cand_pool/global/*.jsonl"}]}, {"config_name": "instructions", "data_files": [{"split": "instructions", "path": "instructions/*.jsonl"}]}, {"config_name": "qrels", "data_files": [{"split": "train", "path": "qrels/train/*.txt"}, {"split": "val", "path": "qrels/val/*.txt"}, {"split": "test", "path": "qrels/test/*.txt"}]}]} | 2024-01-22T00:58:05+00:00 | [
"2311.17136"
] | [
"en"
] | TAGS
#task_categories-text-retrieval #task_categories-text-to-image #task_categories-image-to-text #task_categories-visual-question-answering #language-English #license-mit #arxiv-2311.17136 #region-us
|
### UniIR: Training and Benchmarking Universal Multimodal Information Retrievers
Homepage | Paper | arXiv | GitHub
## News
- [2023-12-21]: Our M-BEIR Benchmark is now available for use.
## Dataset Summary
M-BEIR, the Multimodal BEnchmark for Instructed Retrieval, is a comprehensive large-scale retrieval benchmark designed to train and evaluate unified multimodal retrieval models (UniIR models).
The M-BEIR benchmark comprises eight multimodal retrieval tasks and ten datasets from a variety of domains and sources.
Each task is accompanied by human-authored instructions, encompassing 1.5 million queries and a pool of 5.6 million retrieval candidates in total.
## Dataset Structure Overview
The M-BEIR dataset is structured into five primary components: Query Data, Candidate Pool, Instructions, Qrels, and Images.
### Query Data
Below is the directory structure for the query data:
'train': Contains all the training data from 8 different datasets formatted in the M-BEIR style.
'mbeir_union_up_train.jsonl': This file is the default training data for in-batch contrastive training specifically designed for UniIR models.
It aggregates all the data from the train directory and datasets with relatively smaller sizes have been upsampled to balance the training process.
'val': Contains separate files for validation queries, organized by task.
'test': Contains separate files for test queries, organized by task.
Every M-BEIR query instance has at least one positive candidate data and possibly no negative candidate data
Each line in a Query Data file represents a unique query. The structure of each query JSON object is as follows::
### Candidate Pool
The Candidate Pool contains potential matching documents for the queries.
#### M-BEIR_5.6M
Within the global directory, the default retrieval setting requires models to retrieve positive candidates from a heterogeneous pool encompassing various modalities and domains.
The M-BEIR's global candidate pool, comprising 5.6 million candidates, includes the retrieval corpus from all tasks and datasets.
#### M-BEIR_local
Within the local directory, we provide dataset-task-specific pool as M-BEIR_local. Dataset-task-specific pool contains homogeneous candidates that originate from by the original dataset.
Below is the directory structure for the candidate pool:
The structure of each candidate JSON object in cand_pool file is as follows::
### Instructions
'query_instructions.tsv' contains human-authorized instructions within the UniIR framework. Each task is accompanied by four human-authored instructions. For detailed usage, please refer to GitHub Repo.
### Qrels
Within the 'qrels' directory, you will find qrels for both the validation and test sets. These files serve the purpose of evaluating UniIR models. For detailed information, please refer to GitHub Repo.
## How to Use
### Downloading the M-BEIR Dataset
Clone the M-BEIR repo from the current Page.
Ensure that Git LFS (Large File Storage) is installed on your system, as it will download the required data files.
### Decompressing M-BEIR Images
After downloading, you will need to decompress the image files. Follow these steps in your terminal:
Now, you are ready to use the M-BEIR benchmark.
### Dataloader and Evaluation Pipeline
We offer a dedicated dataloader and evaluation pipeline for the M-BEIR benchmark. Please refer to GitHub Repo for detailed information.
## Citation
Please cite our paper if you use our data, model or code.
| [
"### UniIR: Training and Benchmarking Universal Multimodal Information Retrievers\n Homepage | Paper | arXiv | GitHub",
"## News\n\n- [2023-12-21]: Our M-BEIR Benchmark is now available for use.",
"## Dataset Summary\n\nM-BEIR, the Multimodal BEnchmark for Instructed Retrieval, is a comprehensive large-scale retrieval benchmark designed to train and evaluate unified multimodal retrieval models (UniIR models).\nThe M-BEIR benchmark comprises eight multimodal retrieval tasks and ten datasets from a variety of domains and sources. \nEach task is accompanied by human-authored instructions, encompassing 1.5 million queries and a pool of 5.6 million retrieval candidates in total.",
"## Dataset Structure Overview\nThe M-BEIR dataset is structured into five primary components: Query Data, Candidate Pool, Instructions, Qrels, and Images.",
"### Query Data\n\nBelow is the directory structure for the query data:\n\n'train': Contains all the training data from 8 different datasets formatted in the M-BEIR style.\n\n'mbeir_union_up_train.jsonl': This file is the default training data for in-batch contrastive training specifically designed for UniIR models. \nIt aggregates all the data from the train directory and datasets with relatively smaller sizes have been upsampled to balance the training process.\n\n'val': Contains separate files for validation queries, organized by task.\n\n'test': Contains separate files for test queries, organized by task.\n\nEvery M-BEIR query instance has at least one positive candidate data and possibly no negative candidate data\nEach line in a Query Data file represents a unique query. The structure of each query JSON object is as follows::",
"### Candidate Pool\nThe Candidate Pool contains potential matching documents for the queries.",
"#### M-BEIR_5.6M\nWithin the global directory, the default retrieval setting requires models to retrieve positive candidates from a heterogeneous pool encompassing various modalities and domains. \nThe M-BEIR's global candidate pool, comprising 5.6 million candidates, includes the retrieval corpus from all tasks and datasets.",
"#### M-BEIR_local\nWithin the local directory, we provide dataset-task-specific pool as M-BEIR_local. Dataset-task-specific pool contains homogeneous candidates that originate from by the original dataset.\n\nBelow is the directory structure for the candidate pool:\n\nThe structure of each candidate JSON object in cand_pool file is as follows::",
"### Instructions\n'query_instructions.tsv' contains human-authorized instructions within the UniIR framework. Each task is accompanied by four human-authored instructions. For detailed usage, please refer to GitHub Repo.",
"### Qrels\nWithin the 'qrels' directory, you will find qrels for both the validation and test sets. These files serve the purpose of evaluating UniIR models. For detailed information, please refer to GitHub Repo.",
"## How to Use",
"### Downloading the M-BEIR Dataset\nClone the M-BEIR repo from the current Page. \nEnsure that Git LFS (Large File Storage) is installed on your system, as it will download the required data files.",
"### Decompressing M-BEIR Images\nAfter downloading, you will need to decompress the image files. Follow these steps in your terminal:\n\nNow, you are ready to use the M-BEIR benchmark.",
"### Dataloader and Evaluation Pipeline\nWe offer a dedicated dataloader and evaluation pipeline for the M-BEIR benchmark. Please refer to GitHub Repo for detailed information.",
"## Citation\nPlease cite our paper if you use our data, model or code."
] | [
"TAGS\n#task_categories-text-retrieval #task_categories-text-to-image #task_categories-image-to-text #task_categories-visual-question-answering #language-English #license-mit #arxiv-2311.17136 #region-us \n",
"### UniIR: Training and Benchmarking Universal Multimodal Information Retrievers\n Homepage | Paper | arXiv | GitHub",
"## News\n\n- [2023-12-21]: Our M-BEIR Benchmark is now available for use.",
"## Dataset Summary\n\nM-BEIR, the Multimodal BEnchmark for Instructed Retrieval, is a comprehensive large-scale retrieval benchmark designed to train and evaluate unified multimodal retrieval models (UniIR models).\nThe M-BEIR benchmark comprises eight multimodal retrieval tasks and ten datasets from a variety of domains and sources. \nEach task is accompanied by human-authored instructions, encompassing 1.5 million queries and a pool of 5.6 million retrieval candidates in total.",
"## Dataset Structure Overview\nThe M-BEIR dataset is structured into five primary components: Query Data, Candidate Pool, Instructions, Qrels, and Images.",
"### Query Data\n\nBelow is the directory structure for the query data:\n\n'train': Contains all the training data from 8 different datasets formatted in the M-BEIR style.\n\n'mbeir_union_up_train.jsonl': This file is the default training data for in-batch contrastive training specifically designed for UniIR models. \nIt aggregates all the data from the train directory and datasets with relatively smaller sizes have been upsampled to balance the training process.\n\n'val': Contains separate files for validation queries, organized by task.\n\n'test': Contains separate files for test queries, organized by task.\n\nEvery M-BEIR query instance has at least one positive candidate data and possibly no negative candidate data\nEach line in a Query Data file represents a unique query. The structure of each query JSON object is as follows::",
"### Candidate Pool\nThe Candidate Pool contains potential matching documents for the queries.",
"#### M-BEIR_5.6M\nWithin the global directory, the default retrieval setting requires models to retrieve positive candidates from a heterogeneous pool encompassing various modalities and domains. \nThe M-BEIR's global candidate pool, comprising 5.6 million candidates, includes the retrieval corpus from all tasks and datasets.",
"#### M-BEIR_local\nWithin the local directory, we provide dataset-task-specific pool as M-BEIR_local. Dataset-task-specific pool contains homogeneous candidates that originate from by the original dataset.\n\nBelow is the directory structure for the candidate pool:\n\nThe structure of each candidate JSON object in cand_pool file is as follows::",
"### Instructions\n'query_instructions.tsv' contains human-authorized instructions within the UniIR framework. Each task is accompanied by four human-authored instructions. For detailed usage, please refer to GitHub Repo.",
"### Qrels\nWithin the 'qrels' directory, you will find qrels for both the validation and test sets. These files serve the purpose of evaluating UniIR models. For detailed information, please refer to GitHub Repo.",
"## How to Use",
"### Downloading the M-BEIR Dataset\nClone the M-BEIR repo from the current Page. \nEnsure that Git LFS (Large File Storage) is installed on your system, as it will download the required data files.",
"### Decompressing M-BEIR Images\nAfter downloading, you will need to decompress the image files. Follow these steps in your terminal:\n\nNow, you are ready to use the M-BEIR benchmark.",
"### Dataloader and Evaluation Pipeline\nWe offer a dedicated dataloader and evaluation pipeline for the M-BEIR benchmark. Please refer to GitHub Repo for detailed information.",
"## Citation\nPlease cite our paper if you use our data, model or code."
] | [
75,
32,
24,
125,
43,
207,
20,
83,
91,
59,
57,
4,
53,
47,
41,
17
] | [
"passage: TAGS\n#task_categories-text-retrieval #task_categories-text-to-image #task_categories-image-to-text #task_categories-visual-question-answering #language-English #license-mit #arxiv-2311.17136 #region-us \n### UniIR: Training and Benchmarking Universal Multimodal Information Retrievers\n Homepage | Paper | arXiv | GitHub## News\n\n- [2023-12-21]: Our M-BEIR Benchmark is now available for use.## Dataset Summary\n\nM-BEIR, the Multimodal BEnchmark for Instructed Retrieval, is a comprehensive large-scale retrieval benchmark designed to train and evaluate unified multimodal retrieval models (UniIR models).\nThe M-BEIR benchmark comprises eight multimodal retrieval tasks and ten datasets from a variety of domains and sources. \nEach task is accompanied by human-authored instructions, encompassing 1.5 million queries and a pool of 5.6 million retrieval candidates in total.## Dataset Structure Overview\nThe M-BEIR dataset is structured into five primary components: Query Data, Candidate Pool, Instructions, Qrels, and Images.### Query Data\n\nBelow is the directory structure for the query data:\n\n'train': Contains all the training data from 8 different datasets formatted in the M-BEIR style.\n\n'mbeir_union_up_train.jsonl': This file is the default training data for in-batch contrastive training specifically designed for UniIR models. \nIt aggregates all the data from the train directory and datasets with relatively smaller sizes have been upsampled to balance the training process.\n\n'val': Contains separate files for validation queries, organized by task.\n\n'test': Contains separate files for test queries, organized by task.\n\nEvery M-BEIR query instance has at least one positive candidate data and possibly no negative candidate data\nEach line in a Query Data file represents a unique query. The structure of each query JSON object is as follows::"
] |
a3f026a5e5f1df8e06b843f277a5b66043ac15e8 | # Dataset Card for "gibberish-47-15-08"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | iohadrubin/gibberish-47-15-08 | [
"region:us"
] | 2023-12-01T00:20:14+00:00 | {"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "id", "dtype": "int32"}, {"name": "file_loc", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 70234967, "num_examples": 10475}], "download_size": 39895849, "dataset_size": 70234967}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T00:20:19+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "gibberish-47-15-08"
More Information needed | [
"# Dataset Card for \"gibberish-47-15-08\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"gibberish-47-15-08\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"gibberish-47-15-08\"\n\nMore Information needed"
] |
00c0984770e716799cd80bd77d35e2e6782920f0 | # Dataset Card for "not-gibberish-20-56-22"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | iohadrubin/not-gibberish-20-56-22 | [
"region:us"
] | 2023-12-01T00:27:45+00:00 | {"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "id", "dtype": "int32"}, {"name": "file_loc", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 10140566, "num_examples": 6198}], "download_size": 5851350, "dataset_size": 10140566}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T00:27:47+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "not-gibberish-20-56-22"
More Information needed | [
"# Dataset Card for \"not-gibberish-20-56-22\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"not-gibberish-20-56-22\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"not-gibberish-20-56-22\"\n\nMore Information needed"
] |
8730fae94c4bdba5bd4fc7c1d8368a4552ea18c9 | # Dataset Card for "subj"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | awettig/subj | [
"region:us"
] | 2023-12-01T02:46:14+00:00 | {"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "label", "dtype": "int64"}, {"name": "label_text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1231802, "num_examples": 8000}, {"name": "test", "num_bytes": 310282, "num_examples": 2000}], "download_size": 946189, "dataset_size": 1542084}} | 2023-12-01T02:46:18+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "subj"
More Information needed | [
"# Dataset Card for \"subj\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"subj\"\n\nMore Information needed"
] | [
6,
12
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"subj\"\n\nMore Information needed"
] |
263c2610c16cc19b690230e1b21b91fdf608745b | # Dataset Card for "WMT-year-splits"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | KaiNylund/WMT-year-splits | [
"license:cc0-1.0",
"region:us"
] | 2023-12-01T02:55:19+00:00 | {"license": "cc0-1.0", "dataset_info": {"features": [{"name": "text", "dtype": "string"}], "splits": [{"name": "2012_train", "num_bytes": 200226328, "num_examples": 74030}, {"name": "2012_test", "num_bytes": 10013398, "num_examples": 3702}, {"name": "2013_train", "num_bytes": 200208976, "num_examples": 69560}, {"name": "2013_test", "num_bytes": 10010872, "num_examples": 3452}, {"name": "2014_train", "num_bytes": 200195660, "num_examples": 65066}, {"name": "2014_test", "num_bytes": 10009737, "num_examples": 3178}, {"name": "2015_train", "num_bytes": 200191525, "num_examples": 63260}, {"name": "2015_test", "num_bytes": 10013285, "num_examples": 3193}, {"name": "2016_train", "num_bytes": 200182567, "num_examples": 60204}, {"name": "2016_test", "num_bytes": 10009524, "num_examples": 3068}, {"name": "2017_train", "num_bytes": 200161313, "num_examples": 53757}, {"name": "2017_test", "num_bytes": 10009727, "num_examples": 2712}, {"name": "2018_train", "num_bytes": 200168589, "num_examples": 55074}, {"name": "2018_test", "num_bytes": 10008584, "num_examples": 2780}, {"name": "2019_train", "num_bytes": 200186312, "num_examples": 60742}, {"name": "2019_test", "num_bytes": 10015645, "num_examples": 3082}, {"name": "2020_train", "num_bytes": 200181700, "num_examples": 60036}, {"name": "2020_test", "num_bytes": 10009206, "num_examples": 2932}, {"name": "2021_train", "num_bytes": 200186604, "num_examples": 61717}, {"name": "2021_test", "num_bytes": 10021254, "num_examples": 3001}], "download_size": 1325315435, "dataset_size": 2102010806}} | 2024-02-12T23:27:34+00:00 | [] | [] | TAGS
#license-cc0-1.0 #region-us
| # Dataset Card for "WMT-year-splits"
More Information needed | [
"# Dataset Card for \"WMT-year-splits\"\n\nMore Information needed"
] | [
"TAGS\n#license-cc0-1.0 #region-us \n",
"# Dataset Card for \"WMT-year-splits\"\n\nMore Information needed"
] | [
14,
18
] | [
"passage: TAGS\n#license-cc0-1.0 #region-us \n# Dataset Card for \"WMT-year-splits\"\n\nMore Information needed"
] |
43decb511677c7cd92e2774d274a2cc4ffcd9ed9 | # Dataset Card for "fashion_image_caption-100-v2"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | aisikoduro/fashion_image_caption-100-v2 | [
"region:us"
] | 2023-12-01T03:11:29+00:00 | {"dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 22820471.0, "num_examples": 100}], "download_size": 22820373, "dataset_size": 22820471.0}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T03:11:40+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "fashion_image_caption-100-v2"
More Information needed | [
"# Dataset Card for \"fashion_image_caption-100-v2\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"fashion_image_caption-100-v2\"\n\nMore Information needed"
] | [
6,
20
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"fashion_image_caption-100-v2\"\n\nMore Information needed"
] |
4c876c7647d9ef4b341554af3c090414d2304cde |
# Multilingual Mathematical Autoformalization
["**Paper**"](https://arxiv.org/abs/2311.03755)
This repository contains parallel mathematical statements:
1. Input: An informal proof in natural language
2. Output: The corresponding formalization in either Lean or Isabelle
This dataset can be used to train models how to formalize mathematical statements into verifiable proofs, a form of machine translation.
## Abstract
Autoformalization is the task of translating natural language materials into machine-verifiable formalisations.
Progress in autoformalization research is hindered by the lack of a sizeable dataset consisting of informal-formal pairs expressing the same essence.
Existing methods tend to circumvent this challenge by manually curating small corpora or using few-shot learning with large language models.
But these methods suffer from data scarcity and formal language acquisition difficulty. In this work, we create MMA,
a large, flexible, multilingual, and multi-domain dataset of informal-formal pairs, by using a language model to translate in the reverse direction,
that is, from formal mathematical statements into corresponding informal ones. Experiments show that language models fine-tuned on MMA produce 16−18%
of statements acceptable with minimal corrections on the miniF2F and ProofNet benchmarks, up from 0% with the base model. We demonstrate that fine-tuning
on multilingual formal data results in more capable autoformalization models even when deployed on monolingual tasks.
### Example:
```
Input:
- Statement in natural language: If "r" is a finite set and "i" is an element of "r", then the result of the function "a" applied to "i" is an element of the multiset range of "a" over "r". Translate the statement in natural language to Isabelle:
Output:
- lemma mset_ran_mem[simp, intro]: "finite r \<Longrightarrow> i\<in>r \<Longrightarrow> a i \<in># mset_ran a r"
```
## External Links:
- [**Official GitHub Repository**](https://github.com/albertqjiang/mma)
- [**Papers With Code**](https://paperswithcode.com/paper/multilingual-mathematical-autoformalization)
- [**Arxiv**](https://arxiv.org/abs/2311.03755)
## Citation
```
@misc{jiang2023multilingual,
title={Multilingual Mathematical Autoformalization},
author={Albert Q. Jiang and Wenda Li and Mateja Jamnik},
year={2023},
eprint={2311.03755},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
| casey-martin/multilingual-mathematical-autoformalization | [
"task_categories:translation",
"task_categories:text-generation",
"size_categories:100K<n<1M",
"language:en",
"license:apache-2.0",
"mathematics",
"autoformalization",
"lean",
"isabelle",
"arxiv:2311.03755",
"region:us"
] | 2023-12-01T03:57:06+00:00 | {"language": ["en"], "license": "apache-2.0", "size_categories": ["100K<n<1M"], "task_categories": ["translation", "text-generation"], "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/*_train.jsonl"}, {"split": "val", "path": "data/*_val.jsonl"}, {"split": "test", "path": "data/*_test.jsonl"}]}, {"config_name": "lean", "data_files": [{"split": "train", "path": "data/lean_train.jsonl"}, {"split": "val", "path": "data/lean_val.jsonl"}, {"split": "test", "path": "data/lean_test.jsonl"}]}, {"config_name": "isabelle", "data_files": [{"split": "train", "path": "data/isabelle_train.jsonl"}, {"split": "val", "path": "data/isabelle_val.jsonl"}]}], "tags": ["mathematics", "autoformalization", "lean", "isabelle"]} | 2023-12-02T17:41:02+00:00 | [
"2311.03755"
] | [
"en"
] | TAGS
#task_categories-translation #task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #mathematics #autoformalization #lean #isabelle #arxiv-2311.03755 #region-us
|
# Multilingual Mathematical Autoformalization
"Paper"
This repository contains parallel mathematical statements:
1. Input: An informal proof in natural language
2. Output: The corresponding formalization in either Lean or Isabelle
This dataset can be used to train models how to formalize mathematical statements into verifiable proofs, a form of machine translation.
## Abstract
Autoformalization is the task of translating natural language materials into machine-verifiable formalisations.
Progress in autoformalization research is hindered by the lack of a sizeable dataset consisting of informal-formal pairs expressing the same essence.
Existing methods tend to circumvent this challenge by manually curating small corpora or using few-shot learning with large language models.
But these methods suffer from data scarcity and formal language acquisition difficulty. In this work, we create MMA,
a large, flexible, multilingual, and multi-domain dataset of informal-formal pairs, by using a language model to translate in the reverse direction,
that is, from formal mathematical statements into corresponding informal ones. Experiments show that language models fine-tuned on MMA produce 16−18%
of statements acceptable with minimal corrections on the miniF2F and ProofNet benchmarks, up from 0% with the base model. We demonstrate that fine-tuning
on multilingual formal data results in more capable autoformalization models even when deployed on monolingual tasks.
### Example:
## External Links:
- Official GitHub Repository
- Papers With Code
- Arxiv
| [
"# Multilingual Mathematical Autoformalization\n\"Paper\"\n\nThis repository contains parallel mathematical statements:\n 1. Input: An informal proof in natural language\n 2. Output: The corresponding formalization in either Lean or Isabelle\n\nThis dataset can be used to train models how to formalize mathematical statements into verifiable proofs, a form of machine translation.",
"## Abstract\nAutoformalization is the task of translating natural language materials into machine-verifiable formalisations. \nProgress in autoformalization research is hindered by the lack of a sizeable dataset consisting of informal-formal pairs expressing the same essence. \nExisting methods tend to circumvent this challenge by manually curating small corpora or using few-shot learning with large language models. \nBut these methods suffer from data scarcity and formal language acquisition difficulty. In this work, we create MMA,\na large, flexible, multilingual, and multi-domain dataset of informal-formal pairs, by using a language model to translate in the reverse direction,\nthat is, from formal mathematical statements into corresponding informal ones. Experiments show that language models fine-tuned on MMA produce 16−18%\nof statements acceptable with minimal corrections on the miniF2F and ProofNet benchmarks, up from 0% with the base model. We demonstrate that fine-tuning\non multilingual formal data results in more capable autoformalization models even when deployed on monolingual tasks.",
"### Example:",
"## External Links:\n - Official GitHub Repository\n - Papers With Code\n - Arxiv"
] | [
"TAGS\n#task_categories-translation #task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #mathematics #autoformalization #lean #isabelle #arxiv-2311.03755 #region-us \n",
"# Multilingual Mathematical Autoformalization\n\"Paper\"\n\nThis repository contains parallel mathematical statements:\n 1. Input: An informal proof in natural language\n 2. Output: The corresponding formalization in either Lean or Isabelle\n\nThis dataset can be used to train models how to formalize mathematical statements into verifiable proofs, a form of machine translation.",
"## Abstract\nAutoformalization is the task of translating natural language materials into machine-verifiable formalisations. \nProgress in autoformalization research is hindered by the lack of a sizeable dataset consisting of informal-formal pairs expressing the same essence. \nExisting methods tend to circumvent this challenge by manually curating small corpora or using few-shot learning with large language models. \nBut these methods suffer from data scarcity and formal language acquisition difficulty. In this work, we create MMA,\na large, flexible, multilingual, and multi-domain dataset of informal-formal pairs, by using a language model to translate in the reverse direction,\nthat is, from formal mathematical statements into corresponding informal ones. Experiments show that language models fine-tuned on MMA produce 16−18%\nof statements acceptable with minimal corrections on the miniF2F and ProofNet benchmarks, up from 0% with the base model. We demonstrate that fine-tuning\non multilingual formal data results in more capable autoformalization models even when deployed on monolingual tasks.",
"### Example:",
"## External Links:\n - Official GitHub Repository\n - Papers With Code\n - Arxiv"
] | [
73,
85,
253,
5,
22
] | [
"passage: TAGS\n#task_categories-translation #task_categories-text-generation #size_categories-100K<n<1M #language-English #license-apache-2.0 #mathematics #autoformalization #lean #isabelle #arxiv-2311.03755 #region-us \n# Multilingual Mathematical Autoformalization\n\"Paper\"\n\nThis repository contains parallel mathematical statements:\n 1. Input: An informal proof in natural language\n 2. Output: The corresponding formalization in either Lean or Isabelle\n\nThis dataset can be used to train models how to formalize mathematical statements into verifiable proofs, a form of machine translation.## Abstract\nAutoformalization is the task of translating natural language materials into machine-verifiable formalisations. \nProgress in autoformalization research is hindered by the lack of a sizeable dataset consisting of informal-formal pairs expressing the same essence. \nExisting methods tend to circumvent this challenge by manually curating small corpora or using few-shot learning with large language models. \nBut these methods suffer from data scarcity and formal language acquisition difficulty. In this work, we create MMA,\na large, flexible, multilingual, and multi-domain dataset of informal-formal pairs, by using a language model to translate in the reverse direction,\nthat is, from formal mathematical statements into corresponding informal ones. Experiments show that language models fine-tuned on MMA produce 16−18%\nof statements acceptable with minimal corrections on the miniF2F and ProofNet benchmarks, up from 0% with the base model. We demonstrate that fine-tuning\non multilingual formal data results in more capable autoformalization models even when deployed on monolingual tasks.### Example:## External Links:\n - Official GitHub Repository\n - Papers With Code\n - Arxiv"
] |
b9e92fa5ef5a69910505709233ae5104038d0947 |
- Original Dataset: [Anthropic/hh-rlhf](https://huggingface.co/datasets/Anthropic/hh-rlhf)
- Translation by using [maywell/Synatra-7B-v0.3-Translation](https://huggingface.co/maywell/Synatra-7B-v0.3-Translation)
- Translating in progress... | heegyu/hh-rlhf-ko | [
"language:ko",
"license:mit",
"region:us"
] | 2023-12-01T04:55:40+00:00 | {"language": ["ko"], "license": "mit"} | 2023-12-24T14:35:56+00:00 | [] | [
"ko"
] | TAGS
#language-Korean #license-mit #region-us
|
- Original Dataset: Anthropic/hh-rlhf
- Translation by using maywell/Synatra-7B-v0.3-Translation
- Translating in progress... | [] | [
"TAGS\n#language-Korean #license-mit #region-us \n"
] | [
16
] | [
"passage: TAGS\n#language-Korean #license-mit #region-us \n"
] |
4b32a779ee7e21ce321517d94860d8da94b6e6a1 |
- Original Dataset: [PKU-Alignment/PKU-SafeRLHF](https://huggingface.co/datasets/PKU-Alignment/PKU-SafeRLHF)
- Translation by using [maywell/Synatra-7B-v0.3-Translation](https://huggingface.co/maywell/Synatra-7B-v0.3-Translation)
- Translating in progress... | heegyu/PKU-SafeRLHF-ko | [
"language:ko",
"license:cc-by-nc-4.0",
"region:us"
] | 2023-12-01T05:02:17+00:00 | {"language": ["ko"], "license": "cc-by-nc-4.0"} | 2023-12-31T05:26:05+00:00 | [] | [
"ko"
] | TAGS
#language-Korean #license-cc-by-nc-4.0 #region-us
|
- Original Dataset: PKU-Alignment/PKU-SafeRLHF
- Translation by using maywell/Synatra-7B-v0.3-Translation
- Translating in progress... | [] | [
"TAGS\n#language-Korean #license-cc-by-nc-4.0 #region-us \n"
] | [
22
] | [
"passage: TAGS\n#language-Korean #license-cc-by-nc-4.0 #region-us \n"
] |
ab7e225e475539c8bd15b222e05e4821324bd33d | # Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases
This dataset contains the required data for running the Visual Search Experiment for our NeurIPS paper: [[Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases][neurips]]
Usage instructions can be found on this [GitHub Repo](https://github.com/kreimanlab/VisualSearchAsymmetry)
<br>
# Citations
Shashi Kant Gupta, Mengmi Zhang, Chia-Chien Wu, Jeremy M. Wolfe, & Gabriel Kreiman (2021). Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases. In Thirty-Fifth Conference on Neural Information Processing Systems. [[NeurIPS][neurips]] [[arXiv][arxiv]]
[//]: #
[arxiv]: <https://arxiv.org/abs/2106.02953>
[neurips]: <https://proceedings.neurips.cc/paper_files/paper/2021/hash/37f0e884fbad9667e38940169d0a3c95-Abstract.html> | shashikg/visual_search_klab | [
"arxiv:2106.02953",
"region:us"
] | 2023-12-01T05:22:29+00:00 | {} | 2023-12-01T05:36:49+00:00 | [
"2106.02953"
] | [] | TAGS
#arxiv-2106.02953 #region-us
| # Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases
This dataset contains the required data for running the Visual Search Experiment for our NeurIPS paper: [[Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases][neurips]]
Usage instructions can be found on this GitHub Repo
<br>
s
Shashi Kant Gupta, Mengmi Zhang, Chia-Chien Wu, Jeremy M. Wolfe, & Gabriel Kreiman (2021). Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases. In Thirty-Fifth Conference on Neural Information Processing Systems. [[NeurIPS][neurips]] [[arXiv][arxiv]]
[//]: #
[arxiv]: <URL
[neurips]: <URL | [
"# Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases\n\nThis dataset contains the required data for running the Visual Search Experiment for our NeurIPS paper: [[Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases][neurips]]\n\nUsage instructions can be found on this GitHub Repo\n\n\n<br>\n\ns\n\nShashi Kant Gupta, Mengmi Zhang, Chia-Chien Wu, Jeremy M. Wolfe, & Gabriel Kreiman (2021). Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases. In Thirty-Fifth Conference on Neural Information Processing Systems. [[NeurIPS][neurips]] [[arXiv][arxiv]]\n\n[//]: #\n[arxiv]: <URL\n[neurips]: <URL"
] | [
"TAGS\n#arxiv-2106.02953 #region-us \n",
"# Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases\n\nThis dataset contains the required data for running the Visual Search Experiment for our NeurIPS paper: [[Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases][neurips]]\n\nUsage instructions can be found on this GitHub Repo\n\n\n<br>\n\ns\n\nShashi Kant Gupta, Mengmi Zhang, Chia-Chien Wu, Jeremy M. Wolfe, & Gabriel Kreiman (2021). Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases. In Thirty-Fifth Conference on Neural Information Processing Systems. [[NeurIPS][neurips]] [[arXiv][arxiv]]\n\n[//]: #\n[arxiv]: <URL\n[neurips]: <URL"
] | [
15,
199
] | [
"passage: TAGS\n#arxiv-2106.02953 #region-us \n# Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases\n\nThis dataset contains the required data for running the Visual Search Experiment for our NeurIPS paper: [[Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases][neurips]]\n\nUsage instructions can be found on this GitHub Repo\n\n\n<br>\n\ns\n\nShashi Kant Gupta, Mengmi Zhang, Chia-Chien Wu, Jeremy M. Wolfe, & Gabriel Kreiman (2021). Visual Search Asymmetry: Deep Nets and Humans Share Similar Inherent Biases. In Thirty-Fifth Conference on Neural Information Processing Systems. [[NeurIPS][neurips]] [[arXiv][arxiv]]\n\n[//]: #\n[arxiv]: <URL\n[neurips]: <URL"
] |
25ecdc6c13f0b5d5780b695c7da9adea43572aea | # Dataset Card for "capstone_fromgpt_without_gold_v10_all"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Deojoandco/capstone_fromgpt_without_gold_v10_all | [
"region:us"
] | 2023-12-01T05:28:37+00:00 | {"dataset_info": {"features": [{"name": "dialog_id", "dtype": "int64"}, {"name": "dialogue", "dtype": "string"}, {"name": "summary", "dtype": "string"}, {"name": "gold_tags", "dtype": "string"}, {"name": "gpt_success", "dtype": "bool"}, {"name": "gpt_response", "dtype": "string"}, {"name": "gold_tags_tokens_count", "dtype": "int64"}, {"name": "GPT_TAGS_FOUND", "dtype": "bool"}, {"name": "gpt_output_tags", "dtype": "string"}, {"name": "gpt_output_tag_tokens_count", "dtype": "int64"}, {"name": "GPT_MI_FOUND", "dtype": "bool"}, {"name": "gpt_tags_token_count", "dtype": "int64"}, {"name": "gpt_tags", "dtype": "string"}, {"name": "tag_token_count_match", "dtype": "bool"}], "splits": [{"name": "train", "num_bytes": 124032, "num_examples": 76}, {"name": "validation", "num_bytes": 23025, "num_examples": 12}, {"name": "test", "num_bytes": 14556, "num_examples": 12}], "download_size": 82408, "dataset_size": 161613}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}, {"split": "test", "path": "data/test-*"}]}]} | 2023-12-01T05:32:07+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "capstone_fromgpt_without_gold_v10_all"
More Information needed | [
"# Dataset Card for \"capstone_fromgpt_without_gold_v10_all\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"capstone_fromgpt_without_gold_v10_all\"\n\nMore Information needed"
] | [
6,
26
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"capstone_fromgpt_without_gold_v10_all\"\n\nMore Information needed"
] |
a5ee0f31edb011690c724a71aec4bc36922974ff | a test | sauravprashar/test | [
"region:us"
] | 2023-12-01T06:09:08+00:00 | {} | 2023-12-01T06:32:19+00:00 | [] | [] | TAGS
#region-us
| a test | [] | [
"TAGS\n#region-us \n"
] | [
6
] | [
"passage: TAGS\n#region-us \n"
] |
afb3dc72cbd7ca98a137bad44dc8d03fe51af174 | # Dataset Card for "github-code-haskell-function"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | blastwind/github-code-haskell-function | [
"region:us"
] | 2023-12-01T06:14:03+00:00 | {"dataset_info": {"features": [{"name": "repo_name", "dtype": "string"}, {"name": "path", "dtype": "string"}, {"name": "license", "dtype": "string"}, {"name": "full_code", "dtype": "string"}, {"name": "full_size", "dtype": "int64"}, {"name": "uncommented_code", "dtype": "string"}, {"name": "uncommented_size", "dtype": "int64"}, {"name": "function_only_code", "dtype": "string"}, {"name": "function_only_size", "dtype": "int64"}, {"name": "is_commented", "dtype": "bool"}, {"name": "is_signatured", "dtype": "bool"}, {"name": "n_ast_errors", "dtype": "int64"}, {"name": "ast_max_depth", "dtype": "int64"}, {"name": "n_whitespaces", "dtype": "int64"}, {"name": "n_ast_nodes", "dtype": "int64"}, {"name": "n_ast_terminals", "dtype": "int64"}, {"name": "n_ast_nonterminals", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 2129485080, "num_examples": 2287379}, {"name": "valid", "num_bytes": 303779556, "num_examples": 326768}, {"name": "test", "num_bytes": 609858261, "num_examples": 653538}], "download_size": 1588260184, "dataset_size": 3043122897}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "valid", "path": "data/valid-*"}, {"split": "test", "path": "data/test-*"}]}]} | 2023-12-01T06:17:30+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "github-code-haskell-function"
More Information needed | [
"# Dataset Card for \"github-code-haskell-function\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"github-code-haskell-function\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"github-code-haskell-function\"\n\nMore Information needed"
] |
710e3089b840aa3f049d47955f7cd542c91e23b9 | # Dataset Card for "tab-wnut-packed"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | madaanpulkit/tab-wnut-packed | [
"region:us"
] | 2023-12-01T06:18:13+00:00 | {"dataset_info": {"features": [{"name": "text", "dtype": "string"}, {"name": "tokens", "sequence": "string"}, {"name": "token_spans", "sequence": {"sequence": "int64"}}, {"name": "tags", "sequence": {"sequence": {"class_label": {"names": {"0": "0", "1": "B-DIRECT-CODE", "2": "I-DIRECT-CODE", "3": "B-DIRECT-PERSON", "4": "I-DIRECT-PERSON", "5": "B-QUASI-DATETIME", "6": "I-QUASI-DATETIME", "7": "B-QUASI-PERSON", "8": "I-QUASI-PERSON", "9": "B-QUASI-LOC", "10": "I-QUASI-LOC", "11": "B-QUASI-QUANTITY", "12": "I-QUASI-QUANTITY", "13": "B-QUASI-CODE", "14": "I-QUASI-CODE", "15": "B-QUASI-ORG", "16": "I-QUASI-ORG", "17": "B-QUASI-DEM", "18": "I-QUASI-DEM", "19": "B-QUASI-MISC", "20": "I-QUASI-MISC", "21": "B-DIRECT-ORG", "22": "I-DIRECT-ORG", "23": "B-DIRECT-DATETIME", "24": "I-DIRECT-DATETIME", "25": "B-DIRECT-LOC", "26": "I-DIRECT-LOC", "27": "B-DIRECT-MISC", "28": "I-DIRECT-MISC", "29": "B-DIRECT-DEM", "30": "I-DIRECT-DEM"}}}}}, {"name": "doc_id", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 63243483, "num_examples": 1014}, {"name": "dev", "num_bytes": 7896080, "num_examples": 127}, {"name": "test", "num_bytes": 7787686, "num_examples": 127}], "download_size": 14969043, "dataset_size": 78927249}} | 2023-12-01T17:15:10+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "tab-wnut-packed"
More Information needed | [
"# Dataset Card for \"tab-wnut-packed\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"tab-wnut-packed\"\n\nMore Information needed"
] | [
6,
17
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"tab-wnut-packed\"\n\nMore Information needed"
] |
6b0c111b844e4448b3d88e9d7ef751b56199f22f | # Dataset Card for "vt_multiapi_v0"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Nexusflow/VirusTotalMultiple | [
"region:us"
] | 2023-12-01T06:24:16+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "dataset_info": {"features": [{"name": "fncall", "sequence": "string"}, {"name": "generated_question", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 25028, "num_examples": 70}], "download_size": 12622, "dataset_size": 25028}} | 2023-12-01T06:24:16+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "vt_multiapi_v0"
More Information needed | [
"# Dataset Card for \"vt_multiapi_v0\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"vt_multiapi_v0\"\n\nMore Information needed"
] | [
6,
17
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"vt_multiapi_v0\"\n\nMore Information needed"
] |
024f53228805b33d6ae8ac0a42fe313b1c19f2e9 | # Dataset Card for "new_vt_apis"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Nexusflow/VT_MultiAPIs | [
"region:us"
] | 2023-12-01T06:25:23+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "dataset_info": {"features": [{"name": "args_dicts", "list": [{"name": "default", "dtype": "string"}, {"name": "description", "dtype": "string"}, {"name": "name", "dtype": "string"}, {"name": "required", "dtype": "bool"}, {"name": "type", "dtype": "string"}]}, {"name": "api_type", "dtype": "string"}, {"name": "description", "dtype": "string"}, {"name": "name", "dtype": "string"}, {"name": "dataset", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 20764, "num_examples": 29}], "download_size": 14860, "dataset_size": 20764}} | 2023-12-01T06:25:23+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "new_vt_apis"
More Information needed | [
"# Dataset Card for \"new_vt_apis\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"new_vt_apis\"\n\nMore Information needed"
] | [
6,
16
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"new_vt_apis\"\n\nMore Information needed"
] |
8339ee026230b3536961428e53928f2cfd26ec9b | # Dataset Card for "capstone_fromgpt_without_gold_v11_all"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Deojoandco/capstone_fromgpt_without_gold_v11_all | [
"region:us"
] | 2023-12-01T06:49:08+00:00 | {"dataset_info": {"features": [{"name": "dialog_id", "dtype": "int64"}, {"name": "dialogue", "dtype": "string"}, {"name": "summary", "dtype": "string"}, {"name": "gold_tags", "dtype": "string"}, {"name": "gpt_success", "dtype": "bool"}, {"name": "gpt_response", "dtype": "string"}, {"name": "gold_tags_tokens_count", "dtype": "int64"}, {"name": "GPT_TAGS_FOUND", "dtype": "bool"}, {"name": "gpt_output_tags", "dtype": "string"}, {"name": "gpt_output_tag_tokens_count", "dtype": "int64"}, {"name": "GPT_MI_FOUND", "dtype": "bool"}, {"name": "gpt_tags_token_count", "dtype": "int64"}, {"name": "gpt_tags", "dtype": "string"}, {"name": "tag_token_count_match", "dtype": "bool"}, {"name": "precision", "dtype": "float64"}, {"name": "recall", "dtype": "float64"}, {"name": "f1", "dtype": "float64"}, {"name": "accuracy", "dtype": "float64"}], "splits": [{"name": "validation", "num_bytes": 23400, "num_examples": 12}, {"name": "test", "num_bytes": 14700, "num_examples": 12}], "download_size": 45072, "dataset_size": 38100}, "configs": [{"config_name": "default", "data_files": [{"split": "validation", "path": "data/validation-*"}, {"split": "test", "path": "data/test-*"}]}]} | 2023-12-01T06:50:42+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "capstone_fromgpt_without_gold_v11_all"
More Information needed | [
"# Dataset Card for \"capstone_fromgpt_without_gold_v11_all\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"capstone_fromgpt_without_gold_v11_all\"\n\nMore Information needed"
] | [
6,
26
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"capstone_fromgpt_without_gold_v11_all\"\n\nMore Information needed"
] |
164ce1ee34dd1ab35cb8e7cffb213675a4a68e54 | # Text dataset of Nature family journals
## About
- A scientific document dataset was constructed using open-access papers published by Springer Nature (https://www.springernature.com/). We focused on several journals under the Creative Commons License, including Nature Communications, npj Computational Materials, Nature Computational Science, Communications Chemistry, Communications Materials, and Scientific Reports. From these sources, we collected approximately 65,000 papers published between the 2010s and 2023, containing keywords such as chemistry, synthesis, molecule, polymer, material, and device.
- License
- Articles are distributed under the Creative Commons family license
- e.g., CC 4.0, CC BY-ND 4.0, ...
- Please check raw_data/ref_list.json for each license of the papers
- Files
- raw_data folder
- ref_list.json
- Raw text data of articles
- About 65k papers
- Each record has 'License', 'abstract', 'author', 'bib', 'doi', 'info', 'main', 'other', 'title', and 'ref_id' information.
- context_list.json
- List of introduction part of the articles
- formatted_questions.json
- List of automatically generated questions for some introduction texts
- database folder
- target: texts that are related to the evaluation data
- instruct_eng: test questions and answers
- abst_eng: abstract
- dconcn_eng: conclusion
- intro_eng: introduction
- intro_esp_ger_ita: automatically translated introduction
# NOTE: Texts with CC BY-ND license are excluded because distribution of translated ones is not allowed
- irrelevant 1,2: texts that are not related to the evaluation data
- smallDB folder
- Formatted datasets used for our paper
- qa.json: test questions and answer keywords
- context_ig_paraphrase_plus_oa: context text, their style-changed versions, and irrelevant texts included in "irrelevant 1,2"
---
license: cc
language:
- en
--- | kanhatakeyama/nature-family-CC-papers | [
"region:us"
] | 2023-12-01T07:27:37+00:00 | {} | 2024-02-01T06:33:39+00:00 | [] | [] | TAGS
#region-us
| # Text dataset of Nature family journals
## About
- A scientific document dataset was constructed using open-access papers published by Springer Nature (URL We focused on several journals under the Creative Commons License, including Nature Communications, npj Computational Materials, Nature Computational Science, Communications Chemistry, Communications Materials, and Scientific Reports. From these sources, we collected approximately 65,000 papers published between the 2010s and 2023, containing keywords such as chemistry, synthesis, molecule, polymer, material, and device.
- License
- Articles are distributed under the Creative Commons family license
- e.g., CC 4.0, CC BY-ND 4.0, ...
- Please check raw_data/ref_list.json for each license of the papers
- Files
- raw_data folder
- ref_list.json
- Raw text data of articles
- About 65k papers
- Each record has 'License', 'abstract', 'author', 'bib', 'doi', 'info', 'main', 'other', 'title', and 'ref_id' information.
- context_list.json
- List of introduction part of the articles
- formatted_questions.json
- List of automatically generated questions for some introduction texts
- database folder
- target: texts that are related to the evaluation data
- instruct_eng: test questions and answers
- abst_eng: abstract
- dconcn_eng: conclusion
- intro_eng: introduction
- intro_esp_ger_ita: automatically translated introduction
# NOTE: Texts with CC BY-ND license are excluded because distribution of translated ones is not allowed
- irrelevant 1,2: texts that are not related to the evaluation data
- smallDB folder
- Formatted datasets used for our paper
- URL: test questions and answer keywords
- context_ig_paraphrase_plus_oa: context text, their style-changed versions, and irrelevant texts included in "irrelevant 1,2"
---
license: cc
language:
- en
--- | [
"# Text dataset of Nature family journals",
"## About\n- A scientific document dataset was constructed using open-access papers published by Springer Nature (URL We focused on several journals under the Creative Commons License, including Nature Communications, npj Computational Materials, Nature Computational Science, Communications Chemistry, Communications Materials, and Scientific Reports. From these sources, we collected approximately 65,000 papers published between the 2010s and 2023, containing keywords such as chemistry, synthesis, molecule, polymer, material, and device. \n\n\n- License\n - Articles are distributed under the Creative Commons family license\n - e.g., CC 4.0, CC BY-ND 4.0, ...\n - Please check raw_data/ref_list.json for each license of the papers\n \n- Files\n - raw_data folder\n - ref_list.json\n - Raw text data of articles\n - About 65k papers\n - Each record has 'License', 'abstract', 'author', 'bib', 'doi', 'info', 'main', 'other', 'title', and 'ref_id' information.\n - context_list.json\n - List of introduction part of the articles\n - formatted_questions.json\n - List of automatically generated questions for some introduction texts\n \n - database folder\n - target: texts that are related to the evaluation data\n - instruct_eng: test questions and answers\n - abst_eng: abstract\n - dconcn_eng: conclusion\n - intro_eng: introduction\n - intro_esp_ger_ita: automatically translated introduction\n # NOTE: Texts with CC BY-ND license are excluded because distribution of translated ones is not allowed\n - irrelevant 1,2: texts that are not related to the evaluation data\n \n - smallDB folder\n - Formatted datasets used for our paper\n - URL: test questions and answer keywords\n - context_ig_paraphrase_plus_oa: context text, their style-changed versions, and irrelevant texts included in \"irrelevant 1,2\"\n \n---\nlicense: cc\nlanguage:\n- en\n---"
] | [
"TAGS\n#region-us \n",
"# Text dataset of Nature family journals",
"## About\n- A scientific document dataset was constructed using open-access papers published by Springer Nature (URL We focused on several journals under the Creative Commons License, including Nature Communications, npj Computational Materials, Nature Computational Science, Communications Chemistry, Communications Materials, and Scientific Reports. From these sources, we collected approximately 65,000 papers published between the 2010s and 2023, containing keywords such as chemistry, synthesis, molecule, polymer, material, and device. \n\n\n- License\n - Articles are distributed under the Creative Commons family license\n - e.g., CC 4.0, CC BY-ND 4.0, ...\n - Please check raw_data/ref_list.json for each license of the papers\n \n- Files\n - raw_data folder\n - ref_list.json\n - Raw text data of articles\n - About 65k papers\n - Each record has 'License', 'abstract', 'author', 'bib', 'doi', 'info', 'main', 'other', 'title', and 'ref_id' information.\n - context_list.json\n - List of introduction part of the articles\n - formatted_questions.json\n - List of automatically generated questions for some introduction texts\n \n - database folder\n - target: texts that are related to the evaluation data\n - instruct_eng: test questions and answers\n - abst_eng: abstract\n - dconcn_eng: conclusion\n - intro_eng: introduction\n - intro_esp_ger_ita: automatically translated introduction\n # NOTE: Texts with CC BY-ND license are excluded because distribution of translated ones is not allowed\n - irrelevant 1,2: texts that are not related to the evaluation data\n \n - smallDB folder\n - Formatted datasets used for our paper\n - URL: test questions and answer keywords\n - context_ig_paraphrase_plus_oa: context text, their style-changed versions, and irrelevant texts included in \"irrelevant 1,2\"\n \n---\nlicense: cc\nlanguage:\n- en\n---"
] | [
6,
9,
457
] | [
"passage: TAGS\n#region-us \n# Text dataset of Nature family journals## About\n- A scientific document dataset was constructed using open-access papers published by Springer Nature (URL We focused on several journals under the Creative Commons License, including Nature Communications, npj Computational Materials, Nature Computational Science, Communications Chemistry, Communications Materials, and Scientific Reports. From these sources, we collected approximately 65,000 papers published between the 2010s and 2023, containing keywords such as chemistry, synthesis, molecule, polymer, material, and device. \n\n\n- License\n - Articles are distributed under the Creative Commons family license\n - e.g., CC 4.0, CC BY-ND 4.0, ...\n - Please check raw_data/ref_list.json for each license of the papers\n \n- Files\n - raw_data folder\n - ref_list.json\n - Raw text data of articles\n - About 65k papers\n - Each record has 'License', 'abstract', 'author', 'bib', 'doi', 'info', 'main', 'other', 'title', and 'ref_id' information.\n - context_list.json\n - List of introduction part of the articles\n - formatted_questions.json\n - List of automatically generated questions for some introduction texts\n \n - database folder\n - target: texts that are related to the evaluation data\n - instruct_eng: test questions and answers\n - abst_eng: abstract\n - dconcn_eng: conclusion\n - intro_eng: introduction\n - intro_esp_ger_ita: automatically translated introduction\n # NOTE: Texts with CC BY-ND license are excluded because distribution of translated ones is not allowed\n - irrelevant 1,2: texts that are not related to the evaluation data\n \n - smallDB folder\n - Formatted datasets used for our paper\n - URL: test questions and answer keywords\n - context_ig_paraphrase_plus_oa: context text, their style-changed versions, and irrelevant texts included in \"irrelevant 1,2\"\n \n---\nlicense: cc\nlanguage:\n- en\n---"
] |
fdc8fdb69bf401ad6ef890080f7ccff81952ce51 | # QALD-9-plus Dataset Description
QALD-9-plus is the dataset for Knowledge Graph Question Answering (KGQA) based on well-known [QALD-9](https://github.com/ag-sc/QALD/tree/master/9/data).
QALD-9-plus enables to train and test KGQA systems over DBpedia and Wikidata using questions in 9 different languages: English, German, Russian, French, Armenian, Belarusian, Lithuanian, Bashkir, and Ukrainian.
Some of the questions have several alternative writings in particular languages which enables to evaluate the robustness of KGQA systems and train paraphrasing models.
As the questions' translations were provided by native speakers, they are considered as gold standard, therefore, machine translation tools can be trained and evaluated on the dataset.
# Dataset Statistics
| | en | de | fr | ru | uk | lt | be | ba | hy | # questions DBpedia | # questions Wikidata |
|-------|:---:|:---:|:--:|:----:|:---:|:---:|:---:|:---:|:--:|:-----------:|:-----------:|
| Train | 408 | 543 | 260 | 1203 | 447 | 468 | 441 | 284 | 80 | 408 | 371 |
| Test | 150 | 176 | 26 | 348 | 176 | 186 | 155 | 117 | 20 | 150 | 136 |
Given the numbers, it is obvious that some of the languages are covered more than once i.e., there is more than one translation for a particular question.
For example, there are 1203 Russian translations available while only 408 unique questions exist in the training subset (i.e., 2.9 Russian translations per one question).
The availability of such parallel corpora enables the researchers, developers and other dataset users to address the paraphrasing task.
# Evaluation
We used [GERBIL](https://github.com/dice-group/gerbil/) system for the evaluation of the dataset. The detailed information for the experiments is available at the individual link (click the value in the cells).
## Wikidata
### QAnswer
| | en | de | ru | fr |
|-----|----|----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110010001)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180000)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180001)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180002)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110010007)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180006)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180007)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180008)|
### DeepPavlov
| | en | ru |
|-----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110080010)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180003)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110090001)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180009)|
### Platypus
| | en | fr |
|-----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110110004)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180004)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110110006)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112180010)|
## DBpedia
### QAnswer
| | en | de | ru | fr |
|-----|----|----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110120004)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190000)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190001)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190002)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202110130002)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190003)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190004)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190005)|
## Wikidata Original Translations
### QAnswer
| | de | ru | fr |
|-----|----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190006)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190007)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190008)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190009)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190010)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190011)|
### DeepPavlov
| | ru |
|-----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190012)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190014)|
### Platypus
| | fr |
|-----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190013)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190015)|
## DBpedia Original Translations
### QAnswer
| | de | ru | fr |
|-----|----|----|----|
|Test |[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190016)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190017)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190018)|
|Train|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190019)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190020)|[link](http://gerbil-qa.aksw.org/gerbil/experiment?id=202112190021)|
# Cite
```bibtex
@inproceedings{perevalov2022qald9plus,
author={Perevalov, Aleksandr and Diefenbach, Dennis and Usbeck, Ricardo and Both, Andreas},
booktitle={2022 IEEE 16th International Conference on Semantic Computing (ICSC)},
title={QALD-9-plus: A Multilingual Dataset for Question Answering over DBpedia and Wikidata Translated by Native Speakers},
year={2022},
pages={229-234},
doi={10.1109/ICSC52841.2022.00045}
}
```
# Useful Links
data/* ArXiv [link](https://arxiv.org/abs/2202.00120)
data/* Papers with Code: [Paper](https://paperswithcode.com/paper/qald-9-plus-a-multilingual-dataset-for-1), [Dataset](https://paperswithcode.com/dataset/qald-9-plus)
data/* Video presentation on YouTube: https://youtu.be/W1w7CJTV48c
data/* Presentation [slides](https://drive.google.com/file/d/1cDphq4DeSiZr-WBvdwu34rcxQ0aP4q95/view?usp=sharing)
data/* Google Colab [notebook](https://colab.research.google.com/drive/1eWsQoIaeT9_vii1v3PVU04Rms4EoyLAh?usp=sharing)
# Licence [![CC BY 4.0][cc-by-shield]][cc-by]
This work is licensed under a
[Creative Commons Attribution 4.0 International License][cc-by].
[![CC BY 4.0][cc-by-image]][cc-by]
[cc-by]: http://creativecommons.org/licenses/by/4.0/
[cc-by-image]: https://i.creativecommons.org/l/by/4.0/88x31.png
[cc-by-shield]: https://img.shields.io/badge/License-CC%20BY%204.0-lightgrey.svg
# Dataset Metadata
The following table is necessary for this dataset to be indexed by search
engines such as <a href=https://g.co/datasetsearch>Google Dataset Search</a>.
<div itemscope itemtype=http://schema.org/Dataset>
<table>
<tr>
<th>property</th>
<th>value</th>
</tr>
<tr>
<td>name</td>
<td><code itemprop=name>QALD-9-plus: A Multilingual Dataset for Question Answering over DBpedia and Wikidata Translated by Native Speakers</code></td>
</tr>
<tr>
<td>alternateName</td>
<td><code itemprop=alternateName>QALD-9-plus</code></td>
</tr>
<tr>
<td>url</td>
<td><code itemprop=url>https://github.com/Perevalov/qald_9_plus/tree/main/data</code></td>
</tr>
<tr>
<td>description</td>
<td><code itemprop=description>QALD-9-Plus is the dataset for Knowledge Graph Question Answering (KGQA) based on well-known QALD-9.<br/>
QALD-9-Plus enables to train and test KGQA systems over DBpedia and Wikidata using questions in 9 different languages: English, German, Russian, French, Armenian, Belarusian, Lithuanian, Bashkir, and Ukrainian.<br/>
Some of the questions have several alternative writings in particular languages which enables to evaluate the robustness of KGQA systems and train paraphrasing models.<br/>
As the questions' translations were provided by native speakers, they are considered as gold standard, therefore, machine translation tools can be trained and evaluated on the dataset.</code></td>
</tr>
<tr>
<td>license</td>
<td>
<div itemscope itemtype=http://schema.org/CreativeWork itemprop=license>
<table>
<tr>
<th>property</th>
<th>value</th>
</tr>
<tr>
<td>name</td>
<td><code itemprop=name>CC-BY-4.0</code></td>
</tr>
<tr>
<td>url</td>
<td><code itemprop=url>https://creativecommons.org/licenses/by/4.0/</code></td>
</tr>
</table>
</div>
</td>
</tr>
<tr>
<td>citation</td>
<td><code itemprop=citation>Perevalov, Aleksandr, Diefenbach, Diefenback, Usbeck, Ricardo, Both, Andreas: QALD-9-plus: A multilingual dataset for question answering over DBpedia and Wikidata translated by native speakers. In: 2022 IEEE 16th International Conference on Semantic Computing (ICSC). IEEE (2022)</code></td>
</tr>
</table>
</div>
| casey-martin/qald_9_plus | [
"task_categories:table-question-answering",
"task_categories:text2text-generation",
"language:ba",
"language:be",
"language:de",
"language:en",
"language:fr",
"language:hy",
"language:lt",
"language:ru",
"language:uk",
"license:cc-by-4.0",
"semantic web",
"sparql",
"wikidata",
"dbpedia",
"arxiv:2202.00120",
"region:us"
] | 2023-12-01T07:28:09+00:00 | {"language": ["ba", "be", "de", "en", "fr", "hy", "lt", "ru", "uk"], "license": "cc-by-4.0", "task_categories": ["table-question-answering", "text2text-generation"], "pretty_name": "QALD 9+", "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/*_train.parquet"}]}, {"config_name": "ba", "data_files": [{"split": "train", "path": "data/*_ba_train.parquet"}]}, {"config_name": "be", "data_files": [{"split": "train", "path": "data/*_be_train.parquet"}]}, {"config_name": "de", "data_files": [{"split": "train", "path": "data/*_de_train.parquet"}]}, {"config_name": "en", "data_files": [{"split": "train", "path": "data/*_en_train.parquet"}]}, {"config_name": "fr", "data_files": [{"split": "train", "path": "data/*_fr_train.parquet"}]}, {"config_name": "hy", "data_files": [{"split": "train", "path": "data/*_hy_train.parquet"}]}, {"config_name": "lt", "data_files": [{"split": "train", "path": "data/*_lt_train.parquet"}]}, {"config_name": "ru", "data_files": [{"split": "train", "path": "data/*_ru_train.parquet"}]}, {"config_name": "uk", "data_files": [{"split": "train", "path": "data/*_uk_train.parquet"}]}], "tags": ["semantic web", "sparql", "wikidata", "dbpedia"]} | 2023-12-01T07:36:24+00:00 | [
"2202.00120"
] | [
"ba",
"be",
"de",
"en",
"fr",
"hy",
"lt",
"ru",
"uk"
] | TAGS
#task_categories-table-question-answering #task_categories-text2text-generation #language-Bashkir #language-Belarusian #language-German #language-English #language-French #language-Armenian #language-Lithuanian #language-Russian #language-Ukrainian #license-cc-by-4.0 #semantic web #sparql #wikidata #dbpedia #arxiv-2202.00120 #region-us
| QALD-9-plus Dataset Description
===============================
QALD-9-plus is the dataset for Knowledge Graph Question Answering (KGQA) based on well-known QALD-9.
QALD-9-plus enables to train and test KGQA systems over DBpedia and Wikidata using questions in 9 different languages: English, German, Russian, French, Armenian, Belarusian, Lithuanian, Bashkir, and Ukrainian.
Some of the questions have several alternative writings in particular languages which enables to evaluate the robustness of KGQA systems and train paraphrasing models.
As the questions' translations were provided by native speakers, they are considered as gold standard, therefore, machine translation tools can be trained and evaluated on the dataset.
Dataset Statistics
==================
Given the numbers, it is obvious that some of the languages are covered more than once i.e., there is more than one translation for a particular question.
For example, there are 1203 Russian translations available while only 408 unique questions exist in the training subset (i.e., 2.9 Russian translations per one question).
The availability of such parallel corpora enables the researchers, developers and other dataset users to address the paraphrasing task.
Evaluation
==========
We used GERBIL system for the evaluation of the dataset. The detailed information for the experiments is available at the individual link (click the value in the cells).
Wikidata
--------
### QAnswer
### DeepPavlov
en: Test, ru: link
en: Train, ru: link
### Platypus
en: Test, fr: link
en: Train, fr: link
DBpedia
-------
### QAnswer
Wikidata Original Translations
------------------------------
### QAnswer
### DeepPavlov
### Platypus
DBpedia Original Translations
-----------------------------
### QAnswer
Cite
====
Useful Links
============
data/\* ArXiv link
data/\* Papers with Code: Paper, Dataset
data/\* Video presentation on YouTube: URL
data/\* Presentation slides
data/\* Google Colab notebook
Licence [CC BY 4.0](URL)
========================
This work is licensed under a
[Creative Commons Attribution 4.0 International License](URL).
[](URL)
Dataset Metadata
================
The following table is necessary for this dataset to be indexed by search
engines such as [Google Dataset Search](https://g.co/datasetsearch).
| citation | `Perevalov, Aleksandr, Diefenbach, Diefenback, Usbeck, Ricardo, Both, Andreas: QALD-9-plus: A multilingual dataset for question answering over DBpedia and Wikidata translated by native speakers. In: 2022 IEEE 16th International Conference on Semantic Computing (ICSC). IEEE (2022)` |
| [
"### QAnswer",
"### DeepPavlov\n\n\nen: Test, ru: link\nen: Train, ru: link",
"### Platypus\n\n\nen: Test, fr: link\nen: Train, fr: link\n\n\nDBpedia\n-------",
"### QAnswer\n\n\n\nWikidata Original Translations\n------------------------------",
"### QAnswer",
"### DeepPavlov",
"### Platypus\n\n\n\nDBpedia Original Translations\n-----------------------------",
"### QAnswer\n\n\n\nCite\n====\n\n\nUseful Links\n============\n\n\ndata/\\* ArXiv link\ndata/\\* Papers with Code: Paper, Dataset\ndata/\\* Video presentation on YouTube: URL\ndata/\\* Presentation slides\ndata/\\* Google Colab notebook\n\n\nLicence [CC BY 4.0](URL)\n========================\n\n\nThis work is licensed under a\n[Creative Commons Attribution 4.0 International License](URL).\n\n\n[](URL)\n\n\nDataset Metadata\n================\n\n\nThe following table is necessary for this dataset to be indexed by search\nengines such as [Google Dataset Search](https://g.co/datasetsearch).\n\n\n\n\n\n\n| citation | `Perevalov, Aleksandr, Diefenbach, Diefenback, Usbeck, Ricardo, Both, Andreas: QALD-9-plus: A multilingual dataset for question answering over DBpedia and Wikidata translated by native speakers. In: 2022 IEEE 16th International Conference on Semantic Computing (ICSC). IEEE (2022)` |"
] | [
"TAGS\n#task_categories-table-question-answering #task_categories-text2text-generation #language-Bashkir #language-Belarusian #language-German #language-English #language-French #language-Armenian #language-Lithuanian #language-Russian #language-Ukrainian #license-cc-by-4.0 #semantic web #sparql #wikidata #dbpedia #arxiv-2202.00120 #region-us \n",
"### QAnswer",
"### DeepPavlov\n\n\nen: Test, ru: link\nen: Train, ru: link",
"### Platypus\n\n\nen: Test, fr: link\nen: Train, fr: link\n\n\nDBpedia\n-------",
"### QAnswer\n\n\n\nWikidata Original Translations\n------------------------------",
"### QAnswer",
"### DeepPavlov",
"### Platypus\n\n\n\nDBpedia Original Translations\n-----------------------------",
"### QAnswer\n\n\n\nCite\n====\n\n\nUseful Links\n============\n\n\ndata/\\* ArXiv link\ndata/\\* Papers with Code: Paper, Dataset\ndata/\\* Video presentation on YouTube: URL\ndata/\\* Presentation slides\ndata/\\* Google Colab notebook\n\n\nLicence [CC BY 4.0](URL)\n========================\n\n\nThis work is licensed under a\n[Creative Commons Attribution 4.0 International License](URL).\n\n\n[](URL)\n\n\nDataset Metadata\n================\n\n\nThe following table is necessary for this dataset to be indexed by search\nengines such as [Google Dataset Search](https://g.co/datasetsearch).\n\n\n\n\n\n\n| citation | `Perevalov, Aleksandr, Diefenbach, Diefenback, Usbeck, Ricardo, Both, Andreas: QALD-9-plus: A multilingual dataset for question answering over DBpedia and Wikidata translated by native speakers. In: 2022 IEEE 16th International Conference on Semantic Computing (ICSC). IEEE (2022)` |"
] | [
115,
6,
20,
23,
14,
6,
6,
13,
246
] | [
"passage: TAGS\n#task_categories-table-question-answering #task_categories-text2text-generation #language-Bashkir #language-Belarusian #language-German #language-English #language-French #language-Armenian #language-Lithuanian #language-Russian #language-Ukrainian #license-cc-by-4.0 #semantic web #sparql #wikidata #dbpedia #arxiv-2202.00120 #region-us \n### QAnswer### DeepPavlov\n\n\nen: Test, ru: link\nen: Train, ru: link### Platypus\n\n\nen: Test, fr: link\nen: Train, fr: link\n\n\nDBpedia\n-------### QAnswer\n\n\n\nWikidata Original Translations\n------------------------------### QAnswer### DeepPavlov### Platypus\n\n\n\nDBpedia Original Translations\n-----------------------------### QAnswer\n\n\n\nCite\n====\n\n\nUseful Links\n============\n\n\ndata/\\* ArXiv link\ndata/\\* Papers with Code: Paper, Dataset\ndata/\\* Video presentation on YouTube: URL\ndata/\\* Presentation slides\ndata/\\* Google Colab notebook\n\n\nLicence [CC BY 4.0](URL)\n========================\n\n\nThis work is licensed under a\n[Creative Commons Attribution 4.0 International License](URL).\n\n\n[](URL)\n\n\nDataset Metadata\n================\n\n\nThe following table is necessary for this dataset to be indexed by search\nengines such as [Google Dataset Search](https://g.co/datasetsearch).\n\n\n\n\n\n\n| citation | `Perevalov, Aleksandr, Diefenbach, Diefenback, Usbeck, Ricardo, Both, Andreas: QALD-9-plus: A multilingual dataset for question answering over DBpedia and Wikidata translated by native speakers. In: 2022 IEEE 16th International Conference on Semantic Computing (ICSC). IEEE (2022)` |"
] |
de6f76a9c01bf9b213257e74ea72727ec9fa17ac | oasst1-89k-ja , databricks-dolly-15k-ja , hh-rlhf-49k-ja の中から JGLUE（ JcommonsenseQA , MARC-ja , JSQuAD ）の観点で高品質なデータセットに絞り込んだデータセットです。
品質スコアリングの詳細はこちらを参考にして下さい。
https://qiita.com/kunishou/items/efd9f68d6aa86d56dc73
データセットの構成とライセンスは以下の通りになります。
|dataset|num records|liscence|
|:----|:----|:----|
|[oasst1-89k-ja](https://huggingface.co/datasets/kunishou/oasst1-89k-ja)|4,204|Apache 2.0|
|[databricks-dolly-15k-ja](https://huggingface.co/datasets/kunishou/databricks-dolly-15k-ja)|282|CC-BY-SA-3.0|
|[hh-rlhf-49k-ja](https://huggingface.co/datasets/kunishou/hh-rlhf-49k-ja)|989|[MIT](https://github.com/anthropics/hh-rlhf/blob/master/LICENSE)|
※ ライセンスはオリジナルのデータセットのライセンスに従います。 | kunishou/jp-effective-instructions | [
"language:ja",
"license:other",
"region:us"
] | 2023-12-01T07:36:25+00:00 | {"language": ["ja"], "license": "other", "license_name": "mixed-liscence", "license_link": "LICENSE"} | 2023-12-01T07:52:41+00:00 | [] | [
"ja"
] | TAGS
#language-Japanese #license-other #region-us
| oasst1-89k-ja , databricks-dolly-15k-ja , hh-rlhf-49k-ja の中から JGLUE（ JcommonsenseQA , MARC-ja , JSQuAD ）の観点で高品質なデータセットに絞り込んだデータセットです。
品質スコアリングの詳細はこちらを参考にして下さい。
URL
データセットの構成とライセンスは以下の通りになります。
※ ライセンスはオリジナルのデータセットのライセンスに従います。
| [] | [
"TAGS\n#language-Japanese #license-other #region-us \n"
] | [
17
] | [
"passage: TAGS\n#language-Japanese #license-other #region-us \n"
] |
2f6bd9527c3606e0c57d697a4054204963d7a390 | # Dataset Card for "safety-utcustom-train-v1.0"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | sam1120/safety-utcustom-train-v1.0 | [
"region:us"
] | 2023-12-01T08:33:46+00:00 | {"dataset_info": {"features": [{"name": "name", "dtype": "string"}, {"name": "pixel_values", "dtype": "image"}, {"name": "labels", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 142272068.0, "num_examples": 50}], "download_size": 43505313, "dataset_size": 142272068.0}} | 2023-12-01T08:39:29+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "safety-utcustom-train-v1.0"
More Information needed | [
"# Dataset Card for \"safety-utcustom-train-v1.0\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"safety-utcustom-train-v1.0\"\n\nMore Information needed"
] | [
6,
22
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"safety-utcustom-train-v1.0\"\n\nMore Information needed"
] |
d45ccd131825c3d03b74b32c1012ae2066e1eaca | # Dataset Card for "safety-utcustom-EVAL"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | sam1120/safety-utcustom-EVAL | [
"region:us"
] | 2023-12-01T08:34:32+00:00 | {"dataset_info": {"features": [{"name": "name", "dtype": "string"}, {"name": "pixel_values", "dtype": "image"}, {"name": "labels", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 139241854.0, "num_examples": 50}], "download_size": 40270282, "dataset_size": 139241854.0}} | 2023-12-01T08:41:52+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "safety-utcustom-EVAL"
More Information needed | [
"# Dataset Card for \"safety-utcustom-EVAL\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"safety-utcustom-EVAL\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"safety-utcustom-EVAL\"\n\nMore Information needed"
] |
a1ff219b865bdfb3aa0ed5b48368dd7b04872fbc | # Dataset Card for "safety-utcustom-TEST"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | sam1120/safety-utcustom-TEST | [
"region:us"
] | 2023-12-01T08:37:08+00:00 | {"dataset_info": {"features": [{"name": "name", "dtype": "string"}, {"name": "pixel_values", "dtype": "image"}, {"name": "labels", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 286143613.0, "num_examples": 101}], "download_size": 86634890, "dataset_size": 286143613.0}} | 2023-12-01T08:38:51+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "safety-utcustom-TEST"
More Information needed | [
"# Dataset Card for \"safety-utcustom-TEST\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"safety-utcustom-TEST\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"safety-utcustom-TEST\"\n\nMore Information needed"
] |
58b99c4c80bbc88065b5fba5d326c9f17e94076b | # Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
This dataset card aims to be a base template for new datasets. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed] | Anavya1/Test | [
"language:en",
"region:us"
] | 2023-12-01T10:11:18+00:00 | {"language": ["en"]} | 2023-12-01T10:47:42+00:00 | [] | [
"en"
] | TAGS
#language-English #region-us
| # Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
## Dataset Details
### Dataset Description
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
| [
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
"TAGS\n#language-English #region-us \n",
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
10,
34,
4,
40,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] | [
"passage: TAGS\n#language-English #region-us \n# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.## Dataset Details### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
1bc7c15d1095ee7c60e8ef924c5f5e2ed3a81717 |
# Contents
## Overview
This dataset comprises prompt/answer pairs related to the curriculum for Norwegian dentistry and dental hygiene students, specifically focusing on the subjects of radiation physics, radiation protection, and radiological technology.
## Data files
- data.csv - the original dataset generated as described below and manually proof read
- syntetic_5k_gemini.csv - an augmentation of the original data generated using Google Gemini Pro
## Data source
The prompt/answer pairs in this dataset were generated using commercially available Large Language Models (LLMs), including OpenAI GPT-4 and Anthropic Claude 2. These pairs were generated based on the analysis of documents provided as input to these LLMs.
Following this, some irrelevant pairs were deleted, some were edited for accuracy and clarity, and all pairs were proofread for errors.
### Source documents
The dataset was created using a variety of source documents, primarily encompassing:
- [Lov om strålevern og bruk av stråling (strålevernloven) (2000)](https://lovdata.no/dokument/SF/forskrift/2016-12-16-1659)
- [Forskrift om strålevern og bruk av stråling (strålevernforskriften) (2016)](https://lovdata.no/dokument/SF/forskrift/2016-12-16-1659)
- [DSA: Medisinsk strålebruk (web page) (2023)](https://dsa.no/medisinsk-stralebruk)
- [DSA: Veileder 14](https://dsa.no/publikasjoner/veileder-14-veileder-om-stralebruk-innen-odontologi/Veileder_14_odontologi.pdf)
- [DSA: StrålevernRapport • 2014:2 Strålebruk i Norge](https://dsa.no/publikasjoner/stralevernrapport-2-2014-stralebruk-i-norge/StralevernRapport_02-2014_Str%C3%A5lebruk%20i%20Norge.pdf)
- [DSA: StrålevernRapport 2015:12 Stråledoser til befolkningen](https://dsa.no/publikasjoner/stralevernrapport-12-2015-straledoser-til-befolkningen/StralevernRapport_12-15_Str%C3%A5ledoser_til_befolkningen-.pdf)
- [DSA: Veileder til forskrift om strålevern og bruk av stråling Veileder Nummer 5 Revidert mai 2023](https://dsa.no/publikasjoner/_/attachment/inline/70e8470f-6c36-46fc-9e97-c27298859d66:22ab78bd659798c58cc3ce55c07dbb9aad9b44a0/Veileder%205_rev-mai2023.pdf)
- [Gerald Torgersen: Strålingsfysikk, strålevern og radiologisk teknologi for tannpleie- og tannlegestudenter (online course) (2023)](https://uio.instructure.com/courses/19845)
- Own teaching material and notes
DSA is the The Norwegian Radiation and Nuclear Safety Authority
# Purpose
The dataset is generated for fine-tuning of open source LLMs.
# Format
The dataset is a UTF-8 formatted ";"-separated csv-file. There are two columns: prompt, prediction
# Warning
The dataset is provided for use on own responsibility. Please give feedback if you find a serious error.
# Todo
- add more relevant prompt/response pairs
- further proofreading and adjustments | geraldOslo/RadProtDataSet | [
"size_categories:1K<n<10K",
"language:no",
"license:cc-by-2.0",
"dentistry",
"physics",
"radiation protection",
"region:us"
] | 2023-12-01T10:27:37+00:00 | {"language": ["no"], "license": "cc-by-2.0", "size_categories": ["1K<n<10K"], "pretty_name": "Question/answer connected to radiation protection in dentistry", "tags": ["dentistry", "physics", "radiation protection"], "configs": [{"config_name": "tab", "data_files": ["data.csv", "syntetic_5k_gemini.csv"], "sep": ";"}]} | 2024-01-24T13:16:03+00:00 | [] | [
"no"
] | TAGS
#size_categories-1K<n<10K #language-Norwegian #license-cc-by-2.0 #dentistry #physics #radiation protection #region-us
|
# Contents
## Overview
This dataset comprises prompt/answer pairs related to the curriculum for Norwegian dentistry and dental hygiene students, specifically focusing on the subjects of radiation physics, radiation protection, and radiological technology.
## Data files
- URL - the original dataset generated as described below and manually proof read
- syntetic_5k_gemini.csv - an augmentation of the original data generated using Google Gemini Pro
## Data source
The prompt/answer pairs in this dataset were generated using commercially available Large Language Models (LLMs), including OpenAI GPT-4 and Anthropic Claude 2. These pairs were generated based on the analysis of documents provided as input to these LLMs.
Following this, some irrelevant pairs were deleted, some were edited for accuracy and clarity, and all pairs were proofread for errors.
### Source documents
The dataset was created using a variety of source documents, primarily encompassing:
- Lov om strålevern og bruk av stråling (strålevernloven) (2000)
- Forskrift om strålevern og bruk av stråling (strålevernforskriften) (2016)
- DSA: Medisinsk strålebruk (web page) (2023)
- DSA: Veileder 14
- DSA: StrålevernRapport • 2014:2 Strålebruk i Norge
- DSA: StrålevernRapport 2015:12 Stråledoser til befolkningen
- DSA: Veileder til forskrift om strålevern og bruk av stråling Veileder Nummer 5 Revidert mai 2023
- Gerald Torgersen: Strålingsfysikk, strålevern og radiologisk teknologi for tannpleie- og tannlegestudenter (online course) (2023)
- Own teaching material and notes
DSA is the The Norwegian Radiation and Nuclear Safety Authority
# Purpose
The dataset is generated for fine-tuning of open source LLMs.
# Format
The dataset is a UTF-8 formatted ";"-separated csv-file. There are two columns: prompt, prediction
# Warning
The dataset is provided for use on own responsibility. Please give feedback if you find a serious error.
# Todo
- add more relevant prompt/response pairs
- further proofreading and adjustments | [
"# Contents",
"## Overview\nThis dataset comprises prompt/answer pairs related to the curriculum for Norwegian dentistry and dental hygiene students, specifically focusing on the subjects of radiation physics, radiation protection, and radiological technology.",
"## Data files\n- URL - the original dataset generated as described below and manually proof read\n- syntetic_5k_gemini.csv - an augmentation of the original data generated using Google Gemini Pro",
"## Data source\nThe prompt/answer pairs in this dataset were generated using commercially available Large Language Models (LLMs), including OpenAI GPT-4 and Anthropic Claude 2. These pairs were generated based on the analysis of documents provided as input to these LLMs.\nFollowing this, some irrelevant pairs were deleted, some were edited for accuracy and clarity, and all pairs were proofread for errors.",
"### Source documents\nThe dataset was created using a variety of source documents, primarily encompassing:\n- Lov om strålevern og bruk av stråling (strålevernloven) (2000)\n- Forskrift om strålevern og bruk av stråling (strålevernforskriften) (2016)\n- DSA: Medisinsk strålebruk (web page) (2023)\n- DSA: Veileder 14\n- DSA: StrålevernRapport • 2014:2 Strålebruk i Norge\n- DSA: StrålevernRapport 2015:12 Stråledoser til befolkningen\n- DSA: Veileder til forskrift om strålevern og bruk av stråling Veileder Nummer 5 Revidert mai 2023\n- Gerald Torgersen: Strålingsfysikk, strålevern og radiologisk teknologi for tannpleie- og tannlegestudenter (online course) (2023)\n- Own teaching material and notes\n\nDSA is the The Norwegian Radiation and Nuclear Safety Authority",
"# Purpose\nThe dataset is generated for fine-tuning of open source LLMs.",
"# Format\nThe dataset is a UTF-8 formatted \";\"-separated csv-file. There are two columns: prompt, prediction",
"# Warning\nThe dataset is provided for use on own responsibility. Please give feedback if you find a serious error.",
"# Todo\n- add more relevant prompt/response pairs\n- further proofreading and adjustments"
] | [
"TAGS\n#size_categories-1K<n<10K #language-Norwegian #license-cc-by-2.0 #dentistry #physics #radiation protection #region-us \n",
"# Contents",
"## Overview\nThis dataset comprises prompt/answer pairs related to the curriculum for Norwegian dentistry and dental hygiene students, specifically focusing on the subjects of radiation physics, radiation protection, and radiological technology.",
"## Data files\n- URL - the original dataset generated as described below and manually proof read\n- syntetic_5k_gemini.csv - an augmentation of the original data generated using Google Gemini Pro",
"## Data source\nThe prompt/answer pairs in this dataset were generated using commercially available Large Language Models (LLMs), including OpenAI GPT-4 and Anthropic Claude 2. These pairs were generated based on the analysis of documents provided as input to these LLMs.\nFollowing this, some irrelevant pairs were deleted, some were edited for accuracy and clarity, and all pairs were proofread for errors.",
"### Source documents\nThe dataset was created using a variety of source documents, primarily encompassing:\n- Lov om strålevern og bruk av stråling (strålevernloven) (2000)\n- Forskrift om strålevern og bruk av stråling (strålevernforskriften) (2016)\n- DSA: Medisinsk strålebruk (web page) (2023)\n- DSA: Veileder 14\n- DSA: StrålevernRapport • 2014:2 Strålebruk i Norge\n- DSA: StrålevernRapport 2015:12 Stråledoser til befolkningen\n- DSA: Veileder til forskrift om strålevern og bruk av stråling Veileder Nummer 5 Revidert mai 2023\n- Gerald Torgersen: Strålingsfysikk, strålevern og radiologisk teknologi for tannpleie- og tannlegestudenter (online course) (2023)\n- Own teaching material and notes\n\nDSA is the The Norwegian Radiation and Nuclear Safety Authority",
"# Purpose\nThe dataset is generated for fine-tuning of open source LLMs.",
"# Format\nThe dataset is a UTF-8 formatted \";\"-separated csv-file. There are two columns: prompt, prediction",
"# Warning\nThe dataset is provided for use on own responsibility. Please give feedback if you find a serious error.",
"# Todo\n- add more relevant prompt/response pairs\n- further proofreading and adjustments"
] | [
45,
3,
52,
48,
101,
208,
21,
35,
24,
21
] | [
"passage: TAGS\n#size_categories-1K<n<10K #language-Norwegian #license-cc-by-2.0 #dentistry #physics #radiation protection #region-us \n# Contents## Overview\nThis dataset comprises prompt/answer pairs related to the curriculum for Norwegian dentistry and dental hygiene students, specifically focusing on the subjects of radiation physics, radiation protection, and radiological technology.## Data files\n- URL - the original dataset generated as described below and manually proof read\n- syntetic_5k_gemini.csv - an augmentation of the original data generated using Google Gemini Pro## Data source\nThe prompt/answer pairs in this dataset were generated using commercially available Large Language Models (LLMs), including OpenAI GPT-4 and Anthropic Claude 2. These pairs were generated based on the analysis of documents provided as input to these LLMs.\nFollowing this, some irrelevant pairs were deleted, some were edited for accuracy and clarity, and all pairs were proofread for errors.### Source documents\nThe dataset was created using a variety of source documents, primarily encompassing:\n- Lov om strålevern og bruk av stråling (strålevernloven) (2000)\n- Forskrift om strålevern og bruk av stråling (strålevernforskriften) (2016)\n- DSA: Medisinsk strålebruk (web page) (2023)\n- DSA: Veileder 14\n- DSA: StrålevernRapport • 2014:2 Strålebruk i Norge\n- DSA: StrålevernRapport 2015:12 Stråledoser til befolkningen\n- DSA: Veileder til forskrift om strålevern og bruk av stråling Veileder Nummer 5 Revidert mai 2023\n- Gerald Torgersen: Strålingsfysikk, strålevern og radiologisk teknologi for tannpleie- og tannlegestudenter (online course) (2023)\n- Own teaching material and notes\n\nDSA is the The Norwegian Radiation and Nuclear Safety Authority# Purpose\nThe dataset is generated for fine-tuning of open source LLMs."
] |
98fb04d30fe134c04fcbdfbcffee9326c5c1c08d | # Dataset Card for "instruct-v3"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | kowndinya23/instruct-v3 | [
"region:us"
] | 2023-12-01T10:39:57+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}]}], "dataset_info": {"features": [{"name": "prompt", "dtype": "string"}, {"name": "response", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 219935609, "num_examples": 56167}, {"name": "validation", "num_bytes": 18163894, "num_examples": 6807}], "download_size": 137420551, "dataset_size": 238099503}} | 2023-12-01T10:40:13+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "instruct-v3"
More Information needed | [
"# Dataset Card for \"instruct-v3\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"instruct-v3\"\n\nMore Information needed"
] | [
6,
15
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"instruct-v3\"\n\nMore Information needed"
] |
14a395a263b380f4d4ba3881586917671e67d0f6 | # Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
This dataset card aims to be a base template for new datasets. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
This dataset is the combination of the datasets listed below:
- BDas/Turkish-Dataset
- turkish_product_reviews
- winvoker/turkish-sentiment-analysis-dataset
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed] | bengisucam/tr_dataset_combined | [
"language:tr",
"license:apache-2.0",
"region:us"
] | 2023-12-01T11:21:09+00:00 | {"language": ["tr"], "license": "apache-2.0", "dataset_info": {"features": [{"name": "Text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 167603259, "num_examples": 824809}], "download_size": 106342453, "dataset_size": 167603259}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T14:33:52+00:00 | [] | [
"tr"
] | TAGS
#language-Turkish #license-apache-2.0 #region-us
| # Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
## Dataset Details
### Dataset Description
This dataset is the combination of the datasets listed below:
- BDas/Turkish-Dataset
- turkish_product_reviews
- winvoker/turkish-sentiment-analysis-dataset
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
| [
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\nThis dataset is the combination of the datasets listed below:\n- BDas/Turkish-Dataset\n- turkish_product_reviews\n- winvoker/turkish-sentiment-analysis-dataset\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
"TAGS\n#language-Turkish #license-apache-2.0 #region-us \n",
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\nThis dataset is the combination of the datasets listed below:\n- BDas/Turkish-Dataset\n- turkish_product_reviews\n- winvoker/turkish-sentiment-analysis-dataset\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
20,
34,
4,
88,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] | [
"passage: TAGS\n#language-Turkish #license-apache-2.0 #region-us \n# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.## Dataset Details### Dataset Description\n\n\nThis dataset is the combination of the datasets listed below:\n- BDas/Turkish-Dataset\n- turkish_product_reviews\n- winvoker/turkish-sentiment-analysis-dataset\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
a2b3a1e371e92847b26c65f351ffa990f1981ea9 |
# Dataset Card for Nectar DPO Pairs
| kashif/nectar_dpo_pairs | [
"size_categories:100K<n<1M",
"language:en",
"license:cc-by-nc-4.0",
"RLHF",
"RLAIF",
"reward model",
"region:us"
] | 2023-12-01T11:51:01+00:00 | {"language": ["en"], "license": "cc-by-nc-4.0", "size_categories": ["100K<n<1M"], "datasets": ["berkeley-nest/Nectar"], "dataset_info": {"features": [{"name": "prompt", "dtype": "string"}, {"name": "chosen", "dtype": "string"}, {"name": "rejected", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 8651355540, "num_examples": 3842034}], "download_size": 911865387, "dataset_size": 8651355540}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "tags": ["RLHF", "RLAIF", "reward model"]} | 2023-12-01T12:14:22+00:00 | [] | [
"en"
] | TAGS
#size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #RLHF #RLAIF #reward model #region-us
|
# Dataset Card for Nectar DPO Pairs
| [
"# Dataset Card for Nectar DPO Pairs"
] | [
"TAGS\n#size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #RLHF #RLAIF #reward model #region-us \n",
"# Dataset Card for Nectar DPO Pairs"
] | [
44,
11
] | [
"passage: TAGS\n#size_categories-100K<n<1M #language-English #license-cc-by-nc-4.0 #RLHF #RLAIF #reward model #region-us \n# Dataset Card for Nectar DPO Pairs"
] |
08d245269a0c80f73f4609cfec97cfdfd0bb4780 | # Dataset Card for "ImpartialNews-GenAI-Dataset"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | cmglmsr/ImpartialNews-GenAI-Dataset | [
"region:us"
] | 2023-12-01T11:51:02+00:00 | {"dataset_info": {"features": [{"name": "instruction", "dtype": "string"}, {"name": "input", "dtype": "string"}, {"name": "output", "dtype": "string"}, {"name": "text", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 3595442, "num_examples": 1432}], "download_size": 1467476, "dataset_size": 3595442}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2023-12-01T11:51:05+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "ImpartialNews-GenAI-Dataset"
More Information needed | [
"# Dataset Card for \"ImpartialNews-GenAI-Dataset\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"ImpartialNews-GenAI-Dataset\"\n\nMore Information needed"
] | [
6,
20
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"ImpartialNews-GenAI-Dataset\"\n\nMore Information needed"
] |
7894462f90e17193bcaee5b6c58f087541812af1 |
This dataset contains:
* All reddit posts (submission) and comments of the subreddit reddit.com/r/debateavegan from Jan 20 2011 till Oct 30 2022.
* A set of prompts extracted LLMs fine-tuning based on the highest scoring top-level comments to each reddit post. | lowdewijk/debateavegan_prompts | [
"license:apache-2.0",
"region:us"
] | 2023-12-01T12:29:07+00:00 | {"license": "apache-2.0"} | 2023-12-01T14:03:54+00:00 | [] | [] | TAGS
#license-apache-2.0 #region-us
|
This dataset contains:
* All reddit posts (submission) and comments of the subreddit URL from Jan 20 2011 till Oct 30 2022.
* A set of prompts extracted LLMs fine-tuning based on the highest scoring top-level comments to each reddit post. | [] | [
"TAGS\n#license-apache-2.0 #region-us \n"
] | [
14
] | [
"passage: TAGS\n#license-apache-2.0 #region-us \n"
] |
43b84bf19eb6e4dc4ca157eb87b0eebe31e33b62 | # Dataset Card for "cityscape_11_classes"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Arham-Imran/cityscape_11_classes | [
"region:us"
] | 2023-12-01T13:50:27+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "val", "path": "data/val-*"}]}], "dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "label", "dtype": "image"}], "splits": [{"name": "train", "num_bytes": 6896961361.3, "num_examples": 2975}, {"name": "val", "num_bytes": 1197986021.0, "num_examples": 500}], "download_size": 8226983719, "dataset_size": 8094947382.3}} | 2023-12-01T15:18:04+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "cityscape_11_classes"
More Information needed | [
"# Dataset Card for \"cityscape_11_classes\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"cityscape_11_classes\"\n\nMore Information needed"
] | [
6,
17
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"cityscape_11_classes\"\n\nMore Information needed"
] |
248d16258a5154e70720066ad0d423d7833d02ad | Euskerazko liburuak osaturiko dataseta. Booktegi webgunetik aterata. | Lam-ia/Euskal-liburu-dataseta | [
"task_categories:text-generation",
"size_categories:10K<n<100K",
"language:eu",
"license:apache-2.0",
"region:us"
] | 2023-12-01T14:21:35+00:00 | {"language": ["eu"], "license": "apache-2.0", "size_categories": ["10K<n<100K"], "task_categories": ["text-generation"], "pretty_name": "Euskal Liburuak"} | 2023-12-01T14:25:41+00:00 | [] | [
"eu"
] | TAGS
#task_categories-text-generation #size_categories-10K<n<100K #language-Basque #license-apache-2.0 #region-us
| Euskerazko liburuak osaturiko dataseta. Booktegi webgunetik aterata. | [] | [
"TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-Basque #license-apache-2.0 #region-us \n"
] | [
42
] | [
"passage: TAGS\n#task_categories-text-generation #size_categories-10K<n<100K #language-Basque #license-apache-2.0 #region-us \n"
] |
7f6e382fd19f00983ab690852234e83dfb0e81d2 |
### Description
Scraped articles from alarab (~145,069 articles)
### Usage
```python
from datasets import load_dataset
ds = load_dataset("iahlt/alarab_articles")
```
### Sample
```json
{'url': 'https://www.alarab.co.il/Article/1038538',
'title': 'وفيات النقب | الحاجة فاطمة البحيري في ذمة الله',
'description': 'انتقلت إلى رحمته تعالى مساء اليوم، السبت، في اللقية الحاجة فاطمة حسن الأسد-البحيري (أم سلمان) عن عمر ناهز 93 عاما. والمرحومة هي أرملة المرحوم علي البحيري.ومن المتوقع أن يتم تشييع جثمانها الطاهر إلى أول منازل الآخرة في مقبرة خربة اللقية صب',
'meta_keywords': ['اخبار اليوم، موقع العرب، اخبار العرب، موقع أخبار ، رياضة ، سياسة ، فن عالمي ، فن عربي ، اقتصاد ، موسيقى ، ترفيه ، ألعاب ، سيارات ، أغاني ، كليبات ، افلام عربية ، صور جميلات العرب ومشاهير العرب'],
'tags': ['حالة الطقس',
'اسعار العملات مقابل الشيكل',
'الطقس',
'حالة الطقس اليوم'],
'public_date': '09/07/22 22:40',
'author': 'ياسر العقبي',
'subtitle': 'وفيات النقب | الحاجة فاطمة البحيري في ذمة الله',
'view': 20,
'main_category': 'أخبار',
'sub_category': 'وفيات',
'city': None,
'text': 'انتقلت إلى رحمته تعالى مساء اليوم، السبت، في اللقية الحاجة فاطمة حسن الأسد-البحيري (أم سلمان) عن عمر ناهز 93 عاما. والمرحومة هي أرملة المرحوم علي البحيري.ومن المتوقع أن يتم تشييع جثمانها الطاهر إلى أول منازل الآخرة في مقبرة خربة اللقية صباح يوم غد الأحد.تقبل التعازي في خيمة العزاء بالقرب من بيت الفقيدة.',
'title_len': 46,
'description_len': 238,
'public_date_len': 14.0,
'subtitle_len': 46.0,
'text_len': 304}
```
### Citation
If you use this dataset, please cite:
```
@InProceedings{iahlt2023alarab_articles,
author = "iahlt",
title = "Arabic Corpus: Alarab",
year = "2023",
publisher = "",
location = "",
}
``` | iahlt/alarab_articles | [
"task_categories:feature-extraction",
"task_categories:fill-mask",
"language:ar",
"region:us"
] | 2023-12-01T14:26:34+00:00 | {"language": ["ar"], "task_categories": ["feature-extraction", "fill-mask"], "dataset_info": {"features": [{"name": "url", "dtype": "string"}, {"name": "title", "dtype": "string"}, {"name": "description", "dtype": "string"}, {"name": "meta_keywords", "sequence": "string"}, {"name": "tags", "sequence": "string"}, {"name": "public_date", "dtype": "string"}, {"name": "author", "dtype": "string"}, {"name": "subtitle", "dtype": "string"}, {"name": "view", "dtype": "int64"}, {"name": "main_category", "dtype": "string"}, {"name": "sub_category", "dtype": "string"}, {"name": "city", "dtype": "string"}, {"name": "text", "dtype": "string"}, {"name": "title_len", "dtype": "int64"}, {"name": "description_len", "dtype": "int64"}, {"name": "public_date_len", "dtype": "float64"}, {"name": "subtitle_len", "dtype": "float64"}, {"name": "text_len", "dtype": "int64"}], "splits": [{"name": "train", "num_bytes": 497537790, "num_examples": 145069}], "download_size": 211536806, "dataset_size": 497537790}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}]} | 2024-02-13T07:25:32+00:00 | [] | [
"ar"
] | TAGS
#task_categories-feature-extraction #task_categories-fill-mask #language-Arabic #region-us
|
### Description
Scraped articles from alarab (~145,069 articles)
### Usage
### Sample
If you use this dataset, please cite:
| [
"### Description\n\nScraped articles from alarab (~145,069 articles)",
"### Usage",
"### Sample\n\n\n\nIf you use this dataset, please cite:"
] | [
"TAGS\n#task_categories-feature-extraction #task_categories-fill-mask #language-Arabic #region-us \n",
"### Description\n\nScraped articles from alarab (~145,069 articles)",
"### Usage",
"### Sample\n\n\n\nIf you use this dataset, please cite:"
] | [
34,
17,
4,
14
] | [
"passage: TAGS\n#task_categories-feature-extraction #task_categories-fill-mask #language-Arabic #region-us \n### Description\n\nScraped articles from alarab (~145,069 articles)### Usage### Sample\n\n\n\nIf you use this dataset, please cite:"
] |
2c8f183f996c73fc9607e089c5a5453df6cf934d | CORPUS 1B: A corpus of President George W. Bush's terrorism-related discourse, for a historical research project.
Period: September 11, 2001 (preferable to starting at the beginning of the Bush presidency, for a clean break between pre-9/11 and post-9/11.
-January 20, 2005 (the end of Bush’s first term).
Search parameters: All documents on the American Presidency Project site within the above timeframe returned through a keyword search ‘terror*’
using the wildcard star to return all variants, terrorism, terrorist, etc. Results were further refined to only those associated with
George W. Bush’s name. Further refined by only those documents tagged as ‘spoken addresses or remarks,’
(to filter out noise from some policy papers, and bureaucratic writs and documents).
Also, for the sake of a coherent voice in the fine-tuning data.
Composition: a total of records 1,275 were returned. (including a range of spoken remarks from state of the union addresses,
to remarks to various communities all around the US, to remarks to the press, remarks following discussions with foreign dignitaries, etc).
Here’s the link to these search results: https://www.presidency.ucsb.edu/advanced-search?field-keywords=terror%2A&field-keywords2=&field-keywords3=&from%5Bdate%5D=09-11-2001&to%5Bdate%5D=01-20-2005&person2=&items_per_page=100&f%5B0%5D=field_docs_person%3A200299&f%5B1%5D=field_docs_category%3A8
Word count pre-cleaning: 2,227,662
Word count post-cleaning: 2,236,541
| GPT-JF/Corpus_1B | [
"region:us"
] | 2023-12-01T14:46:50+00:00 | {} | 2023-12-18T15:35:03+00:00 | [] | [] | TAGS
#region-us
| CORPUS 1B: A corpus of President George W. Bush's terrorism-related discourse, for a historical research project.
Period: September 11, 2001 (preferable to starting at the beginning of the Bush presidency, for a clean break between pre-9/11 and post-9/11.
-January 20, 2005 (the end of Bush’s first term).
Search parameters: All documents on the American Presidency Project site within the above timeframe returned through a keyword search ‘terror*’
using the wildcard star to return all variants, terrorism, terrorist, etc. Results were further refined to only those associated with
George W. Bush’s name. Further refined by only those documents tagged as ‘spoken addresses or remarks,’
(to filter out noise from some policy papers, and bureaucratic writs and documents).
Also, for the sake of a coherent voice in the fine-tuning data.
Composition: a total of records 1,275 were returned. (including a range of spoken remarks from state of the union addresses,
to remarks to various communities all around the US, to remarks to the press, remarks following discussions with foreign dignitaries, etc).
Here’s the link to these search results: URL
Word count pre-cleaning: 2,227,662
Word count post-cleaning: 2,236,541
| [] | [
"TAGS\n#region-us \n"
] | [
6
] | [
"passage: TAGS\n#region-us \n"
] |
6b501f525c9177ce4af24e20276f8e01af4ac5f0 |
## Description
I am a chess expert.
## Prompt
A video channel managed by a renowed grandmaster, Mongoose Carlsun.
The videos are informative, but playful and fun.
| Xenova/ai-tube-my-chess-bot | [
"license:cc-by-nc-sa-4.0",
"ai-tube:Chess Bot",
"region:us"
] | 2023-12-01T14:57:05+00:00 | {"license": "cc-by-nc-sa-4.0", "pretty_name": "My Chess Bot", "tags": ["ai-tube:Chess Bot"]} | 2023-12-01T14:58:26+00:00 | [] | [] | TAGS
#license-cc-by-nc-sa-4.0 #ai-tube-Chess Bot #region-us
|
## Description
I am a chess expert.
## Prompt
A video channel managed by a renowed grandmaster, Mongoose Carlsun.
The videos are informative, but playful and fun.
| [
"## Description\n\nI am a chess expert.",
"## Prompt\n\nA video channel managed by a renowed grandmaster, Mongoose Carlsun.\n\nThe videos are informative, but playful and fun."
] | [
"TAGS\n#license-cc-by-nc-sa-4.0 #ai-tube-Chess Bot #region-us \n",
"## Description\n\nI am a chess expert.",
"## Prompt\n\nA video channel managed by a renowed grandmaster, Mongoose Carlsun.\n\nThe videos are informative, but playful and fun."
] | [
28,
10,
33
] | [
"passage: TAGS\n#license-cc-by-nc-sa-4.0 #ai-tube-Chess Bot #region-us \n## Description\n\nI am a chess expert.## Prompt\n\nA video channel managed by a renowed grandmaster, Mongoose Carlsun.\n\nThe videos are informative, but playful and fun."
] |
77f7e271bf4a92b24fce5119f3486b583ca016ff | # Syntec dataset for information retrieval
This dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.
## Dataset Details
The dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.
It is split into 2 subsets :
- **queries** : it features 100 manually created questions. Each question is mapped to the article that contains the answer.
- **documents** : corresponds to the 90 articles from the collective bargaining
### Usage
```py
import datasets
# Download the documents (corpus)
corpus_raw = datasets.load_dataset("lyon-nlp/mteb-fr-retrieval-syntec-s2p", "documents")
# Download the queries
queries_raw = datasets.load_dataset("lyon-nlp/mteb-fr-retrieval-syntec-s2p", "queries")
```
### Dataset Description
The collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.
The dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.
- **Curated by:** Wikit AI (https://www.wikit.ai/)
- **Language(s) (NLP):** French
- **License:** [More Information Needed]
### Dataset Sources
https://www.syntec.fr/
### Contact
[email protected]
[email protected] | lyon-nlp/mteb-fr-retrieval-syntec-s2p | [
"task_categories:question-answering",
"language:fr",
"region:us"
] | 2023-12-01T15:16:19+00:00 | {"language": ["fr"], "task_categories": ["question-answering"], "pretty_name": "Syntec dataset for information retrieval"} | 2023-12-04T09:19:06+00:00 | [] | [
"fr"
] | TAGS
#task_categories-question-answering #language-French #region-us
| # Syntec dataset for information retrieval
This dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.
## Dataset Details
The dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.
It is split into 2 subsets :
- queries : it features 100 manually created questions. Each question is mapped to the article that contains the answer.
- documents : corresponds to the 90 articles from the collective bargaining
### Usage
### Dataset Description
The collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.
The dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.
- Curated by: Wikit AI (URL
- Language(s) (NLP): French
- License:
### Dataset Sources
URL
### Contact
mathieu@URL
marion@URL | [
"# Syntec dataset for information retrieval\n\nThis dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.",
"## Dataset Details\n\nThe dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.\n\nIt is split into 2 subsets :\n- queries : it features 100 manually created questions. Each question is mapped to the article that contains the answer.\n- documents : corresponds to the 90 articles from the collective bargaining",
"### Usage",
"### Dataset Description\n\nThe collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.\n\nThe dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.\n\n- Curated by: Wikit AI (URL\n- Language(s) (NLP): French\n- License:",
"### Dataset Sources\n\nURL",
"### Contact\n\nmathieu@URL\nmarion@URL"
] | [
"TAGS\n#task_categories-question-answering #language-French #region-us \n",
"# Syntec dataset for information retrieval\n\nThis dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.",
"## Dataset Details\n\nThe dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.\n\nIt is split into 2 subsets :\n- queries : it features 100 manually created questions. Each question is mapped to the article that contains the answer.\n- documents : corresponds to the 90 articles from the collective bargaining",
"### Usage",
"### Dataset Description\n\nThe collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.\n\nThe dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.\n\n- Curated by: Wikit AI (URL\n- Language(s) (NLP): French\n- License:",
"### Dataset Sources\n\nURL",
"### Contact\n\nmathieu@URL\nmarion@URL"
] | [
24,
36,
79,
4,
98,
7,
12
] | [
"passage: TAGS\n#task_categories-question-answering #language-French #region-us \n# Syntec dataset for information retrieval\n\nThis dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.## Dataset Details\n\nThe dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.\n\nIt is split into 2 subsets :\n- queries : it features 100 manually created questions. Each question is mapped to the article that contains the answer.\n- documents : corresponds to the 90 articles from the collective bargaining### Usage### Dataset Description\n\nThe collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.\n\nThe dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.\n\n- Curated by: Wikit AI (URL\n- Language(s) (NLP): French\n- License:### Dataset Sources\n\nURL### Contact\n\nmathieu@URL\nmarion@URL"
] |
f459757b3b01ad5525f2c29d720b78a8791f729f |
## Description
In a galaxy far far away, there was a wholesome Knight and a small creature as companion.
## Prompt
A video channel managed by the famous Space Knight Djin Darin.
The videos are scenery of galaxy, futuristic knights, aliens, planets, spacecrafts.
The humor should be about how the absurd the small creature acts.
The video will be with trips about the stories of the knight and a small creature, life on different planets. | merve/ai-tube-dummy | [
"license:apache-2.0",
"ai-tube:Dummy",
"region:us"
] | 2023-12-01T15:23:57+00:00 | {"license": "apache-2.0", "pretty_name": "AI Tube", "tags": ["ai-tube:Dummy"]} | 2023-12-01T15:56:25+00:00 | [] | [] | TAGS
#license-apache-2.0 #ai-tube-Dummy #region-us
|
## Description
In a galaxy far far away, there was a wholesome Knight and a small creature as companion.
## Prompt
A video channel managed by the famous Space Knight Djin Darin.
The videos are scenery of galaxy, futuristic knights, aliens, planets, spacecrafts.
The humor should be about how the absurd the small creature acts.
The video will be with trips about the stories of the knight and a small creature, life on different planets. | [
"## Description\n\nIn a galaxy far far away, there was a wholesome Knight and a small creature as companion.",
"## Prompt\n\nA video channel managed by the famous Space Knight Djin Darin.\n\nThe videos are scenery of galaxy, futuristic knights, aliens, planets, spacecrafts.\n\nThe humor should be about how the absurd the small creature acts.\n\nThe video will be with trips about the stories of the knight and a small creature, life on different planets."
] | [
"TAGS\n#license-apache-2.0 #ai-tube-Dummy #region-us \n",
"## Description\n\nIn a galaxy far far away, there was a wholesome Knight and a small creature as companion.",
"## Prompt\n\nA video channel managed by the famous Space Knight Djin Darin.\n\nThe videos are scenery of galaxy, futuristic knights, aliens, planets, spacecrafts.\n\nThe humor should be about how the absurd the small creature acts.\n\nThe video will be with trips about the stories of the knight and a small creature, life on different planets."
] | [
22,
22,
81
] | [
"passage: TAGS\n#license-apache-2.0 #ai-tube-Dummy #region-us \n## Description\n\nIn a galaxy far far away, there was a wholesome Knight and a small creature as companion.## Prompt\n\nA video channel managed by the famous Space Knight Djin Darin.\n\nThe videos are scenery of galaxy, futuristic knights, aliens, planets, spacecrafts.\n\nThe humor should be about how the absurd the small creature acts.\n\nThe video will be with trips about the stories of the knight and a small creature, life on different planets."
] |
a23b50e70fbe5241897e41dde7c465b1450944a3 | # Dataset Card for "Impressions"
## Overview
The Impressions dataset is a multimodal benchmark that consists of 4,100 unique annotations and over 1,375 image-caption pairs from the photography domain. Each annotation explores (1) the aesthetic impactfulness of a photograph, (2) image descriptions in which pragmatic inferences are welcome, (3) emotions/thoughts/beliefs that the photograph may inspire, and (4) the aesthetic elements that elicited the expressed impression.
EMNLP 2023 | [Paper](https://arxiv.org/abs/2310.17887)
## Additional Data
The Impressions dataset comes with more information that just image annotations on questions pertaining to *Pragmatic Description*, *Perception*, and *Aesthetic Evaluation*. For annotator personality and demographic metadata, as well as all *Aesthetic Impact* annotations, please see our [git repository](https://github.com/SALT-NLP/Impressions)!
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | SALT-NLP/Impressions | [
"task_categories:image-to-text",
"task_categories:visual-question-answering",
"size_categories:1K<n<10K",
"language:en",
"license:cc-by-sa-4.0",
"art",
"arxiv:2310.17887",
"region:us"
] | 2023-12-01T15:39:53+00:00 | {"language": ["en"], "license": "cc-by-sa-4.0", "size_categories": ["1K<n<10K"], "task_categories": ["image-to-text", "visual-question-answering"], "pretty_name": "Impressions", "dataset_info": {"features": [{"name": "image", "dtype": "image"}, {"name": "AnnotatorId", "dtype": "string"}, {"name": "ImgId", "dtype": "string"}, {"name": "caption", "dtype": "string"}, {"name": "Impact", "dtype": "float64"}, {"name": "image_description", "dtype": "string"}, {"name": "image_impression", "dtype": "string"}, {"name": "image_aesthetic_eval", "dtype": "string"}, {"name": "image_url", "dtype": "string"}, {"name": "source", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 2366953929.024, "num_examples": 1352}], "download_size": 2214475090, "dataset_size": 2366953929.024}, "tags": ["art"]} | 2023-12-01T15:58:34+00:00 | [
"2310.17887"
] | [
"en"
] | TAGS
#task_categories-image-to-text #task_categories-visual-question-answering #size_categories-1K<n<10K #language-English #license-cc-by-sa-4.0 #art #arxiv-2310.17887 #region-us
| # Dataset Card for "Impressions"
## Overview
The Impressions dataset is a multimodal benchmark that consists of 4,100 unique annotations and over 1,375 image-caption pairs from the photography domain. Each annotation explores (1) the aesthetic impactfulness of a photograph, (2) image descriptions in which pragmatic inferences are welcome, (3) emotions/thoughts/beliefs that the photograph may inspire, and (4) the aesthetic elements that elicited the expressed impression.
EMNLP 2023 | Paper
## Additional Data
The Impressions dataset comes with more information that just image annotations on questions pertaining to *Pragmatic Description*, *Perception*, and *Aesthetic Evaluation*. For annotator personality and demographic metadata, as well as all *Aesthetic Impact* annotations, please see our git repository!
More Information needed | [
"# Dataset Card for \"Impressions\"",
"## Overview\n\nThe Impressions dataset is a multimodal benchmark that consists of 4,100 unique annotations and over 1,375 image-caption pairs from the photography domain. Each annotation explores (1) the aesthetic impactfulness of a photograph, (2) image descriptions in which pragmatic inferences are welcome, (3) emotions/thoughts/beliefs that the photograph may inspire, and (4) the aesthetic elements that elicited the expressed impression.\n\nEMNLP 2023 | Paper",
"## Additional Data\n\nThe Impressions dataset comes with more information that just image annotations on questions pertaining to *Pragmatic Description*, *Perception*, and *Aesthetic Evaluation*. For annotator personality and demographic metadata, as well as all *Aesthetic Impact* annotations, please see our git repository! \n\n\nMore Information needed"
] | [
"TAGS\n#task_categories-image-to-text #task_categories-visual-question-answering #size_categories-1K<n<10K #language-English #license-cc-by-sa-4.0 #art #arxiv-2310.17887 #region-us \n",
"# Dataset Card for \"Impressions\"",
"## Overview\n\nThe Impressions dataset is a multimodal benchmark that consists of 4,100 unique annotations and over 1,375 image-caption pairs from the photography domain. Each annotation explores (1) the aesthetic impactfulness of a photograph, (2) image descriptions in which pragmatic inferences are welcome, (3) emotions/thoughts/beliefs that the photograph may inspire, and (4) the aesthetic elements that elicited the expressed impression.\n\nEMNLP 2023 | Paper",
"## Additional Data\n\nThe Impressions dataset comes with more information that just image annotations on questions pertaining to *Pragmatic Description*, *Perception*, and *Aesthetic Evaluation*. For annotator personality and demographic metadata, as well as all *Aesthetic Impact* annotations, please see our git repository! \n\n\nMore Information needed"
] | [
71,
10,
112,
85
] | [
"passage: TAGS\n#task_categories-image-to-text #task_categories-visual-question-answering #size_categories-1K<n<10K #language-English #license-cc-by-sa-4.0 #art #arxiv-2310.17887 #region-us \n# Dataset Card for \"Impressions\"## Overview\n\nThe Impressions dataset is a multimodal benchmark that consists of 4,100 unique annotations and over 1,375 image-caption pairs from the photography domain. Each annotation explores (1) the aesthetic impactfulness of a photograph, (2) image descriptions in which pragmatic inferences are welcome, (3) emotions/thoughts/beliefs that the photograph may inspire, and (4) the aesthetic elements that elicited the expressed impression.\n\nEMNLP 2023 | Paper## Additional Data\n\nThe Impressions dataset comes with more information that just image annotations on questions pertaining to *Pragmatic Description*, *Perception*, and *Aesthetic Evaluation*. For annotator personality and demographic metadata, as well as all *Aesthetic Impact* annotations, please see our git repository! \n\n\nMore Information needed"
] |
68cf25ad9390f90a530e3e72f4762cd8cb6d63cd | # Dataset Card for "capstone_fromgpt_without_gold_v12_all"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | Deojoandco/capstone_fromgpt_without_gold_v12_all | [
"region:us"
] | 2023-12-01T15:47:57+00:00 | {"dataset_info": {"features": [{"name": "dialog_id", "dtype": "int64"}, {"name": "dialogue", "dtype": "string"}, {"name": "summary", "dtype": "string"}, {"name": "gold_tags", "dtype": "string"}, {"name": "gpt_success", "dtype": "bool"}, {"name": "gpt_response", "dtype": "string"}, {"name": "gold_tags_tokens_count", "dtype": "int64"}, {"name": "GPT_TAGS_FOUND", "dtype": "bool"}, {"name": "gpt_output_tags", "dtype": "string"}, {"name": "gpt_output_tag_tokens_count", "dtype": "int64"}, {"name": "GPT_MI_FOUND", "dtype": "bool"}, {"name": "gpt_tags_token_count", "dtype": "int64"}, {"name": "gpt_tags", "dtype": "string"}, {"name": "tag_token_count_match", "dtype": "bool"}, {"name": "precision", "dtype": "float64"}, {"name": "recall", "dtype": "float64"}, {"name": "f1", "dtype": "float64"}, {"name": "accuracy", "dtype": "float64"}], "splits": [{"name": "validation", "num_bytes": 23408, "num_examples": 12}], "download_size": 25882, "dataset_size": 23408}, "configs": [{"config_name": "default", "data_files": [{"split": "validation", "path": "data/validation-*"}]}]} | 2023-12-01T15:49:46+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "capstone_fromgpt_without_gold_v12_all"
More Information needed | [
"# Dataset Card for \"capstone_fromgpt_without_gold_v12_all\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"capstone_fromgpt_without_gold_v12_all\"\n\nMore Information needed"
] | [
6,
26
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"capstone_fromgpt_without_gold_v12_all\"\n\nMore Information needed"
] |
cef928e8b1381e66abce384de322cdc5c43c32ad | # Dataset Card for "autotrain-data-flan_test"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | healthcorum/autotrain-data-flan_test | [
"region:us"
] | 2023-12-01T16:01:17+00:00 | {"dataset_info": {"features": [{"name": "autotrain_text", "dtype": "string"}, {"name": "autotrain_label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 9436691, "num_examples": 7998}, {"name": "validation", "num_bytes": 2359962, "num_examples": 2000}], "download_size": 4112546, "dataset_size": 11796653}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "validation", "path": "data/validation-*"}]}]} | 2023-12-01T16:01:19+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "autotrain-data-flan_test"
More Information needed | [
"# Dataset Card for \"autotrain-data-flan_test\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"autotrain-data-flan_test\"\n\nMore Information needed"
] | [
6,
19
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"autotrain-data-flan_test\"\n\nMore Information needed"
] |
da2bccb35790a24297f56300cd1ce755994914c9 |
# [Music-Driven Group Choreography (CVPR 2023)](https://openaccess.thecvf.com/content/CVPR2023/papers/Le_Music-Driven_Group_Choreography_CVPR_2023_paper.pdf)
### *[Nhat Le](https://minhnhatvt.github.io/), [Thang Pham](https://phamtrongthang123.github.io/), [Tuong Do](https://scholar.google.com/citations?user=qCcSKkMAAAAJ&hl=en), [Erman Tjiputra](https://sg.linkedin.com/in/erman-tjiputra), [Quang D. Tran](https://scholar.google.com/citations?user=DbAThEgAAAAJ&hl=en), [Anh Nguyen](https://cgi.csc.liv.ac.uk/~anguyen/)*
### [[Project Page](https://aioz-ai.github.io/AIOZ-GDANCE/)] [[Paper](https://openaccess.thecvf.com/content/CVPR2023/papers/Le_Music-Driven_Group_Choreography_CVPR_2023_paper.pdf)]
*<center> We demonstrate the AIOZ-GDANCE dataset with in-the-wild videos, music audio, and 3D group dance motion. </center>*
## Abstract
> Music-driven choreography is a challenging problem with a wide variety of industrial applications. Recently, many methods have been proposed to synthesize dance motions from music for a single dancer. However, generating dance motion for a group remains an open problem. In this paper, we present GDANCE, a new large-scale dataset for music-driven group dance generation. Unlike existing datasets that only support single dance, our new dataset contains group dance videos, hence supporting the study of group choreography. We propose a semi-autonomous labeling method with humans in the loop to obtain the 3D ground truth for our dataset. The proposed dataset consists of 16.7 hours of paired music and 3D motion from in-the-wild videos, covering 7 dance styles and 16 music genres. We show that naively applying single dance generation technique to creating group dance motion may lead to unsatisfactory results, such as inconsistent movements and collisions between dancers. Based on our new dataset, we propose a new method that takes an input music sequence and a set of 3D positions of dancers to efficiently produce multiple group-coherent choreographies. We propose new evaluation metrics for measuring group dance quality and perform intensive experiments to demonstrate the effectiveness of our method. Our code and dataset will be released to facilitate future research on group dance generation.
## Table of Contents
1. [AIOZ-GDANCE Dataset](#aioz-gdance-dataset)
2. [Visualizing](#visualizing)
3. [Prerequisites](#prerequisites)
4. [Usage](#usage)
## AIOZ-GDANCE Dataset
**[Download]** The dataset can be downloaded at [Data](https://vision.aioz.io/f/430eb9d90552480e8b4e/?dl=1)
**[Updated]** The music and dance labels are now available at [Labels](https://vision.aioz.io/f/bef3ae93990e4e43addf/?dl=1)
The data directory is organized as follows:
- **split_sequence_names.txt**:
- a txt file containing seperate sequence names in the data (each sequence should have unique name or id)
- **musics**:
- contains raw music .wav file of each sequence with the corresponding name. The music frames are aligned with the motion frames.
- **motions_smpl**:
- contains the motion file of each sequence with the corresponding name, the motion is provided in .pkl file format.
- Each data dictionary mainly includes the following items:
- `'smpl_poses': shape[num_persons x num_frames x 72]`: the motions contain 72-D vector pose sequences in SMPL pose format (24 joints).
- ``'root_trans': shape[num_persons x num_frames x 3]``: sequences of root translation.
Here is an example python script to read the motion file
```python
import pickle
import numpy as np
data = pickle.load(open("sequence_name.pkl","rb"))
print(data.keys())
smpl_poses = data['smpl_poses']
smpl_trans = data['root_trans']
# ... may utilize the pose by using SMPL forward function: https://github.com/vchoutas/smplx
```
## Visualizing
We provide demo code for loading and visualizing the motions.
### Prerequisites
First, you need to download the [SMPL model](https://smpl.is.tue.mpg.de/) (v1.0.0) and rename the model files for visualization. The directory structure of the data is expected to be:
The directory structure of the data is expected to be:
```
<DATA_DIR>
├── motions_smpl/
├── musics/
└── split_sequence_names.txt
<SMPL_DIR>
├── SMPL_MALE.pkl
└── SMPL_FEMALE.pkl
```
Then run this to install the necessary packages
```
pip install scipy torch smplx chumpy vedo trimesh
pip install numpy==1.23.0
```
### Usage
#### Visualize the SMPL joints
The following command will first calculate the SMPL joint locations (joint rotations and root translation) and then plot on the 3D figure in realtime.
``` bash
python vis_smpl_kpt.py \
--data_dir <DATA_DIR>/motions_smpl \
--smpl_path <SMPL_DIR>/SMPL_FEMALE.PKL \
--sequence_name sequence_name.pkl
```
#### Visualize the SMPL Mesh
The following command will calculate the SMPL meshes and visualize in 3D.
``` bash
python vis_smpl_mesh.py \
--data_dir <DATA_DIR>/motions_smpl \
--smpl_path <SMPL_DIR>/SMPL_FEMALE.PKL \
--sequence_name sequence_name.pkl
```
## TODO
- [x] ~~**Dataset**~~
- [ ] **Baseline model & training**: TBD
## Citation
```
@inproceedings{aiozGdance,
author = {Le, Nhat and Pham, Thang and Do, Tuong and Tjiputra, Erman and Tran, Quang D. and Nguyen, Anh},
title = {Music-Driven Group Choreography},
journal = {CVPR},
year = {2023},
}
```
## License
Software Copyright License for non-commercial scientific research purposes.
Please read carefully the following [terms and conditions](LICENSE) and any accompanying
documentation before you download and/or use AIOZ-GDANCE data, model and
software, (the "Data & Software"), including 3D meshes, images, videos,
textures, software, scripts, and animations. By downloading and/or using the
Data & Software (including downloading, cloning, installing, and any other use
of the corresponding github repository), you acknowledge that you have read
these [terms and conditions](LICENSE), understand them, and agree to be bound by them. If
you do not agree with these [terms and conditions](LICENSE), you must not download and/or
use the Data & Software. Any infringement of the terms of this agreement will
automatically terminate your rights under this [License](LICENSE).
## Acknowledgement
This repo used visualization code from [AIST++](https://github.com/google/aistplusplus_api/tree/main) | aiozai/AIOZ-GDANCE | [
"language:en",
"license:other",
"region:us"
] | 2023-12-01T16:37:55+00:00 | {"language": ["en"], "license": "other", "license_name": "aioz-license", "license_link": "LICENSE"} | 2023-12-04T02:32:54+00:00 | [] | [
"en"
] | TAGS
#language-English #license-other #region-us
|
# Music-Driven Group Choreography (CVPR 2023)
### *Nhat Le, Thang Pham, Tuong Do, Erman Tjiputra, Quang D. Tran, Anh Nguyen*
### [Project Page] [Paper]
- musics:
- contains raw music .wav file of each sequence with the corresponding name. The music frames are aligned with the motion frames.
- motions_smpl:
- contains the motion file of each sequence with the corresponding name, the motion is provided in .pkl file format.
- Each data dictionary mainly includes the following items:
- ''smpl_poses': shape[num_persons x num_frames x 72]': the motions contain 72-D vector pose sequences in SMPL pose format (24 joints).
- '''root_trans': shape[num_persons x num_frames x 3]'': sequences of root translation.
Here is an example python script to read the motion file
!Figure 4
!Figure 5
## Visualizing
We provide demo code for loading and visualizing the motions.
### Prerequisites
First, you need to download the SMPL model (v1.0.0) and rename the model files for visualization. The directory structure of the data is expected to be:
The directory structure of the data is expected to be:
Then run this to install the necessary packages
### Usage
#### Visualize the SMPL joints
The following command will first calculate the SMPL joint locations (joint rotations and root translation) and then plot on the 3D figure in realtime.
#### Visualize the SMPL Mesh
The following command will calculate the SMPL meshes and visualize in 3D.
## TODO
- [x] ~~Dataset~~
- [ ] Baseline model & training: TBD
## License
Software Copyright License for non-commercial scientific research purposes.
Please read carefully the following terms and conditions and any accompanying
documentation before you download and/or use AIOZ-GDANCE data, model and
software, (the "Data & Software"), including 3D meshes, images, videos,
textures, software, scripts, and animations. By downloading and/or using the
Data & Software (including downloading, cloning, installing, and any other use
of the corresponding github repository), you acknowledge that you have read
these terms and conditions, understand them, and agree to be bound by them. If
you do not agree with these terms and conditions, you must not download and/or
use the Data & Software. Any infringement of the terms of this agreement will
automatically terminate your rights under this License.
## Acknowledgement
This repo used visualization code from AIST++ | [
"# Music-Driven Group Choreography (CVPR 2023)",
"### *Nhat Le, Thang Pham, Tuong Do, Erman Tjiputra, Quang D. Tran, Anh Nguyen*",
"### [Project Page] [Paper]\n\n\n![](URL We demonstrate the AIOZ-GDANCE dataset with in-the-wild videos, music audio, and 3D group dance motion. </center>*",
"## Abstract\n> Music-driven choreography is a challenging problem with a wide variety of industrial applications. Recently, many methods have been proposed to synthesize dance motions from music for a single dancer. However, generating dance motion for a group remains an open problem. In this paper, we present GDANCE, a new large-scale dataset for music-driven group dance generation. Unlike existing datasets that only support single dance, our new dataset contains group dance videos, hence supporting the study of group choreography. We propose a semi-autonomous labeling method with humans in the loop to obtain the 3D ground truth for our dataset. The proposed dataset consists of 16.7 hours of paired music and 3D motion from in-the-wild videos, covering 7 dance styles and 16 music genres. We show that naively applying single dance generation technique to creating group dance motion may lead to unsatisfactory results, such as inconsistent movements and collisions between dancers. Based on our new dataset, we propose a new method that takes an input music sequence and a set of 3D positions of dancers to efficiently produce multiple group-coherent choreographies. We propose new evaluation metrics for measuring group dance quality and perform intensive experiments to demonstrate the effectiveness of our method. Our code and dataset will be released to facilitate future research on group dance generation.",
"## Table of Contents\n1. AIOZ-GDANCE Dataset\n2. Visualizing\n3. Prerequisites\n4. Usage",
"## AIOZ-GDANCE Dataset\n[Download] The dataset can be downloaded at Data\n\n[Updated] The music and dance labels are now available at Labels\n\nThe data directory is organized as follows:\n- split_sequence_names.txt:\n - a txt file containing seperate sequence names in the data (each sequence should have unique name or id)\n- musics:\n - contains raw music .wav file of each sequence with the corresponding name. The music frames are aligned with the motion frames.\n- motions_smpl:\n - contains the motion file of each sequence with the corresponding name, the motion is provided in .pkl file format.\n - Each data dictionary mainly includes the following items:\n - ''smpl_poses': shape[num_persons x num_frames x 72]': the motions contain 72-D vector pose sequences in SMPL pose format (24 joints).\n - '''root_trans': shape[num_persons x num_frames x 3]'': sequences of root translation.\n\nHere is an example python script to read the motion file\n\n!Figure 4 \n\n!Figure 5",
"## Visualizing\n\n\nWe provide demo code for loading and visualizing the motions.",
"### Prerequisites\nFirst, you need to download the SMPL model (v1.0.0) and rename the model files for visualization. The directory structure of the data is expected to be:\n\nThe directory structure of the data is expected to be:\n\n\nThen run this to install the necessary packages",
"### Usage",
"#### Visualize the SMPL joints\nThe following command will first calculate the SMPL joint locations (joint rotations and root translation) and then plot on the 3D figure in realtime.",
"#### Visualize the SMPL Mesh\nThe following command will calculate the SMPL meshes and visualize in 3D.",
"## TODO\n- [x] ~~Dataset~~\n- [ ] Baseline model & training: TBD",
"## License\nSoftware Copyright License for non-commercial scientific research purposes.\nPlease read carefully the following terms and conditions and any accompanying\ndocumentation before you download and/or use AIOZ-GDANCE data, model and\nsoftware, (the \"Data & Software\"), including 3D meshes, images, videos,\ntextures, software, scripts, and animations. By downloading and/or using the\nData & Software (including downloading, cloning, installing, and any other use\nof the corresponding github repository), you acknowledge that you have read\nthese terms and conditions, understand them, and agree to be bound by them. If\nyou do not agree with these terms and conditions, you must not download and/or\nuse the Data & Software. Any infringement of the terms of this agreement will\nautomatically terminate your rights under this License.",
"## Acknowledgement\nThis repo used visualization code from AIST++"
] | [
"TAGS\n#language-English #license-other #region-us \n",
"# Music-Driven Group Choreography (CVPR 2023)",
"### *Nhat Le, Thang Pham, Tuong Do, Erman Tjiputra, Quang D. Tran, Anh Nguyen*",
"### [Project Page] [Paper]\n\n\n![](URL We demonstrate the AIOZ-GDANCE dataset with in-the-wild videos, music audio, and 3D group dance motion. </center>*",
"## Abstract\n> Music-driven choreography is a challenging problem with a wide variety of industrial applications. Recently, many methods have been proposed to synthesize dance motions from music for a single dancer. However, generating dance motion for a group remains an open problem. In this paper, we present GDANCE, a new large-scale dataset for music-driven group dance generation. Unlike existing datasets that only support single dance, our new dataset contains group dance videos, hence supporting the study of group choreography. We propose a semi-autonomous labeling method with humans in the loop to obtain the 3D ground truth for our dataset. The proposed dataset consists of 16.7 hours of paired music and 3D motion from in-the-wild videos, covering 7 dance styles and 16 music genres. We show that naively applying single dance generation technique to creating group dance motion may lead to unsatisfactory results, such as inconsistent movements and collisions between dancers. Based on our new dataset, we propose a new method that takes an input music sequence and a set of 3D positions of dancers to efficiently produce multiple group-coherent choreographies. We propose new evaluation metrics for measuring group dance quality and perform intensive experiments to demonstrate the effectiveness of our method. Our code and dataset will be released to facilitate future research on group dance generation.",
"## Table of Contents\n1. AIOZ-GDANCE Dataset\n2. Visualizing\n3. Prerequisites\n4. Usage",
"## AIOZ-GDANCE Dataset\n[Download] The dataset can be downloaded at Data\n\n[Updated] The music and dance labels are now available at Labels\n\nThe data directory is organized as follows:\n- split_sequence_names.txt:\n - a txt file containing seperate sequence names in the data (each sequence should have unique name or id)\n- musics:\n - contains raw music .wav file of each sequence with the corresponding name. The music frames are aligned with the motion frames.\n- motions_smpl:\n - contains the motion file of each sequence with the corresponding name, the motion is provided in .pkl file format.\n - Each data dictionary mainly includes the following items:\n - ''smpl_poses': shape[num_persons x num_frames x 72]': the motions contain 72-D vector pose sequences in SMPL pose format (24 joints).\n - '''root_trans': shape[num_persons x num_frames x 3]'': sequences of root translation.\n\nHere is an example python script to read the motion file\n\n!Figure 4 \n\n!Figure 5",
"## Visualizing\n\n\nWe provide demo code for loading and visualizing the motions.",
"### Prerequisites\nFirst, you need to download the SMPL model (v1.0.0) and rename the model files for visualization. The directory structure of the data is expected to be:\n\nThe directory structure of the data is expected to be:\n\n\nThen run this to install the necessary packages",
"### Usage",
"#### Visualize the SMPL joints\nThe following command will first calculate the SMPL joint locations (joint rotations and root translation) and then plot on the 3D figure in realtime.",
"#### Visualize the SMPL Mesh\nThe following command will calculate the SMPL meshes and visualize in 3D.",
"## TODO\n- [x] ~~Dataset~~\n- [ ] Baseline model & training: TBD",
"## License\nSoftware Copyright License for non-commercial scientific research purposes.\nPlease read carefully the following terms and conditions and any accompanying\ndocumentation before you download and/or use AIOZ-GDANCE data, model and\nsoftware, (the \"Data & Software\"), including 3D meshes, images, videos,\ntextures, software, scripts, and animations. By downloading and/or using the\nData & Software (including downloading, cloning, installing, and any other use\nof the corresponding github repository), you acknowledge that you have read\nthese terms and conditions, understand them, and agree to be bound by them. If\nyou do not agree with these terms and conditions, you must not download and/or\nuse the Data & Software. Any infringement of the terms of this agreement will\nautomatically terminate your rights under this License.",
"## Acknowledgement\nThis repo used visualization code from AIST++"
] | [
15,
14,
31,
49,
325,
24,
276,
17,
64,
4,
43,
27,
23,
180,
15
] | [
"passage: TAGS\n#language-English #license-other #region-us \n# Music-Driven Group Choreography (CVPR 2023)### *Nhat Le, Thang Pham, Tuong Do, Erman Tjiputra, Quang D. Tran, Anh Nguyen*### [Project Page] [Paper]\n\n\n![](URL We demonstrate the AIOZ-GDANCE dataset with in-the-wild videos, music audio, and 3D group dance motion. </center>*## Abstract\n> Music-driven choreography is a challenging problem with a wide variety of industrial applications. Recently, many methods have been proposed to synthesize dance motions from music for a single dancer. However, generating dance motion for a group remains an open problem. In this paper, we present GDANCE, a new large-scale dataset for music-driven group dance generation. Unlike existing datasets that only support single dance, our new dataset contains group dance videos, hence supporting the study of group choreography. We propose a semi-autonomous labeling method with humans in the loop to obtain the 3D ground truth for our dataset. The proposed dataset consists of 16.7 hours of paired music and 3D motion from in-the-wild videos, covering 7 dance styles and 16 music genres. We show that naively applying single dance generation technique to creating group dance motion may lead to unsatisfactory results, such as inconsistent movements and collisions between dancers. Based on our new dataset, we propose a new method that takes an input music sequence and a set of 3D positions of dancers to efficiently produce multiple group-coherent choreographies. We propose new evaluation metrics for measuring group dance quality and perform intensive experiments to demonstrate the effectiveness of our method. Our code and dataset will be released to facilitate future research on group dance generation.## Table of Contents\n1. AIOZ-GDANCE Dataset\n2. Visualizing\n3. Prerequisites\n4. Usage"
] |
5dc3c1449066c6b4efae9f42402262cf06e1e32c | # [doc] image dataset 9
This dataset contains 2 jpeg files in the `red` directory and 2 jpeg files in the `green` directory. The YAML header in this README disables the automatic "label" column. | datasets-examples/doc-image-9 | [
"size_categories:n<1K",
"region:us"
] | 2023-12-01T17:26:41+00:00 | {"size_categories": ["n<1K"], "configs": [{"config_name": "default", "drop_labels": true}]} | 2023-12-01T17:29:48+00:00 | [] | [] | TAGS
#size_categories-n<1K #region-us
| # [doc] image dataset 9
This dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory. The YAML header in this README disables the automatic "label" column. | [
"# [doc] image dataset 9\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory. The YAML header in this README disables the automatic \"label\" column."
] | [
"TAGS\n#size_categories-n<1K #region-us \n",
"# [doc] image dataset 9\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory. The YAML header in this README disables the automatic \"label\" column."
] | [
16,
57
] | [
"passage: TAGS\n#size_categories-n<1K #region-us \n# [doc] image dataset 9\n\nThis dataset contains 2 jpeg files in the 'red' directory and 2 jpeg files in the 'green' directory. The YAML header in this README disables the automatic \"label\" column."
] |
b1906050b3d3bd0320eae5fa2fe1600ab7365fb7 |
This is a code specific dataset. It contains two-turn questions and answers. The first question/answer pair comes from glaiveai/glaive-code-assistant-v2. The second question (follow-up) was generated and answered by Claude-2. | migtissera/Tess-Coder-v1.0 | [
"license:apache-2.0",
"region:us"
] | 2023-12-01T17:49:09+00:00 | {"license": "apache-2.0"} | 2023-12-16T18:58:01+00:00 | [] | [] | TAGS
#license-apache-2.0 #region-us
|
This is a code specific dataset. It contains two-turn questions and answers. The first question/answer pair comes from glaiveai/glaive-code-assistant-v2. The second question (follow-up) was generated and answered by Claude-2. | [] | [
"TAGS\n#license-apache-2.0 #region-us \n"
] | [
14
] | [
"passage: TAGS\n#license-apache-2.0 #region-us \n"
] |
ec97b77e96b8c92da0371dc3725f1c23fc641d21 | # Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic
**Homepage:** [https://www.stratosphereips.org/datasets-iot23](https://www.stratosphereips.org/datasets-iot23)
This dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices [(Comparative Analysis of IoT Botnet Datasets)](https://doi.org/10.53070/bbd.1173687).
The selection of the subset was determined by [Aqeel Ahmed on Kaggle](https://www.kaggle.com/datasets/engraqeel/iot23preprocesseddata) and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of conn.log.labelled files.
This dataset only notes if the data is Malcious or Benign. The original dataset labels the type of malcious traffic aswell. This means this processing of the dataset is only suited for binary classification.
# Feature information:
All features originate from the [Zeek](https://docs.zeek.org/en/master/scripts/base/protocols/conn/main.zeek.html#type-Conn::Info) processing performed by the dataset creators. [See notes here for caviats for each column](https://docs.zeek.org/en/master/scripts/base/protocols/conn/main.zeek.html#type-Conn::Info).
<details>
<summary>Expand for feature names, descriptions, and datatypes</summary>
Name: id.orig_p
Description: The originator’s port number.
Data type: int64 - uint64 in original
Name: id.resp_p
Description: The responder’s port number.
Data type: int64 - uint64 in original
Name: proto
Description: The transport layer protocol of the connection.
Data type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset
Name: service
Description: An identification of an application protocol being sent over the connection.
Data type: optional string
Name: duration
Description: How long the connection lasted.
Data type: optional float64 - time interval
Name: orig_bytes
Description: The number of payload bytes the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_bytes
Description:The number of payload bytes the responder sent.
Data type: optional int64 - uint64 in original
Name: conn_state
Description: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH)
Data type: optional string
Name: missed_bytes
Description: Indicates the number of bytes missed in content gaps, which is representative of packet loss.
Data type: optional int64 - uint64 in original. default = 0
Name: history
Description: Records the state history of connections as a string of letters.
Data type: optional string
Name: orig_pkts
Description: Number of packets that the originator sent.
Data type: optional int64 - uint64 in original
Name: orig_ip_bytes
Description: Number of IP level bytes that the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_pkts
Description: Number of packets that the responder sent.
Data type: optional int64 - uint64 in original
Name: resp_ip_bytes
Description: Number of IP level bytes that the responder sent.
Data type: optional int64 - uint64 in original
Name: label
Description: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types
Data type: string - enum(Malicious, Benign)
NOTE: ts, uid, id.orig_h, id.resp_h have been removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h) using this dataset, as that can lead to over fitting to dataset specific times and addresses.
Further local_orig, local_resp have been removed as they are null in all rows, so they are useless for training.
</details>
## Citation
If you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. http://doi.org/10.5281/zenodo.4743746” | yashika0998/iot-23-preprocessed | [
"task_categories:question-answering",
"task_categories:tabular-classification",
"language:en",
"code",
"region:us"
] | 2023-12-01T18:02:50+00:00 | {"language": ["en"], "task_categories": ["question-answering", "tabular-classification"], "pretty_name": "d", "dataset_info": {"features": [{"name": "id.orig_p", "dtype": "int64"}, {"name": "id.resp_p", "dtype": "int64"}, {"name": "proto", "dtype": "string"}, {"name": "service", "dtype": "string"}, {"name": "duration", "dtype": "float64"}, {"name": "orig_bytes", "dtype": "int64"}, {"name": "resp_bytes", "dtype": "int64"}, {"name": "conn_state", "dtype": "string"}, {"name": "missed_bytes", "dtype": "int64"}, {"name": "history", "dtype": "string"}, {"name": "orig_pkts", "dtype": "int64"}, {"name": "orig_ip_bytes", "dtype": "int64"}, {"name": "resp_pkts", "dtype": "int64"}, {"name": "resp_ip_bytes", "dtype": "int64"}, {"name": "label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 93994789, "num_examples": 819024}], "download_size": 11805369, "dataset_size": 93994789}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "tags": ["code"]} | 2023-12-01T18:27:06+00:00 | [] | [
"en"
] | TAGS
#task_categories-question-answering #task_categories-tabular-classification #language-English #code #region-us
| # Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic
Homepage: URL
This dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets).
The selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.
This dataset only notes if the data is Malcious or Benign. The original dataset labels the type of malcious traffic aswell. This means this processing of the dataset is only suited for binary classification.
# Feature information:
All features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column.
<details>
<summary>Expand for feature names, descriptions, and datatypes</summary>
Name: id.orig_p
Description: The originator’s port number.
Data type: int64 - uint64 in original
Name: id.resp_p
Description: The responder’s port number.
Data type: int64 - uint64 in original
Name: proto
Description: The transport layer protocol of the connection.
Data type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset
Name: service
Description: An identification of an application protocol being sent over the connection.
Data type: optional string
Name: duration
Description: How long the connection lasted.
Data type: optional float64 - time interval
Name: orig_bytes
Description: The number of payload bytes the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_bytes
Description:The number of payload bytes the responder sent.
Data type: optional int64 - uint64 in original
Name: conn_state
Description: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH)
Data type: optional string
Name: missed_bytes
Description: Indicates the number of bytes missed in content gaps, which is representative of packet loss.
Data type: optional int64 - uint64 in original. default = 0
Name: history
Description: Records the state history of connections as a string of letters.
Data type: optional string
Name: orig_pkts
Description: Number of packets that the originator sent.
Data type: optional int64 - uint64 in original
Name: orig_ip_bytes
Description: Number of IP level bytes that the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_pkts
Description: Number of packets that the responder sent.
Data type: optional int64 - uint64 in original
Name: resp_ip_bytes
Description: Number of IP level bytes that the responder sent.
Data type: optional int64 - uint64 in original
Name: label
Description: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types
Data type: string - enum(Malicious, Benign)
NOTE: ts, uid, id.orig_h, id.resp_h have been removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h) using this dataset, as that can lead to over fitting to dataset specific times and addresses.
Further local_orig, local_resp have been removed as they are null in all rows, so they are useless for training.
</details>
If you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL” | [
"# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.\n\nThis dataset only notes if the data is Malcious or Benign. The original dataset labels the type of malcious traffic aswell. This means this processing of the dataset is only suited for binary classification.",
"# Feature information:\n\nAll features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column. \n<details>\n <summary>Expand for feature names, descriptions, and datatypes</summary>\n\nName: id.orig_p \nDescription: The originator’s port number. \nData type: int64 - uint64 in original \n\nName: id.resp_p \nDescription: The responder’s port number. \nData type: int64 - uint64 in original \n\nName: proto \nDescription: The transport layer protocol of the connection. \nData type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset \n\nName: service \nDescription: An identification of an application protocol being sent over the connection. \nData type: optional string \n\nName: duration \nDescription: How long the connection lasted. \nData type: optional float64 - time interval \n\nName: orig_bytes \nDescription: The number of payload bytes the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_bytes \nDescription:The number of payload bytes the responder sent. \nData type: optional int64 - uint64 in original \n\nName: conn_state \nDescription: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH) \nData type: optional string \n\nName: missed_bytes \nDescription: Indicates the number of bytes missed in content gaps, which is representative of packet loss. \nData type: optional int64 - uint64 in original. default = 0\n\nName: history \nDescription: Records the state history of connections as a string of letters. \nData type: optional string \n\nName: orig_pkts \nDescription: Number of packets that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: orig_ip_bytes \nDescription: Number of IP level bytes that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_pkts \nDescription: Number of packets that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: resp_ip_bytes \nDescription: Number of IP level bytes that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: label \nDescription: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types \nData type: string - enum(Malicious, Benign)\n\nNOTE: ts, uid, id.orig_h, id.resp_h have been removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h) using this dataset, as that can lead to over fitting to dataset specific times and addresses. \nFurther local_orig, local_resp have been removed as they are null in all rows, so they are useless for training.\n</details>\n\nIf you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL”"
] | [
"TAGS\n#task_categories-question-answering #task_categories-tabular-classification #language-English #code #region-us \n",
"# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.\n\nThis dataset only notes if the data is Malcious or Benign. The original dataset labels the type of malcious traffic aswell. This means this processing of the dataset is only suited for binary classification.",
"# Feature information:\n\nAll features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column. \n<details>\n <summary>Expand for feature names, descriptions, and datatypes</summary>\n\nName: id.orig_p \nDescription: The originator’s port number. \nData type: int64 - uint64 in original \n\nName: id.resp_p \nDescription: The responder’s port number. \nData type: int64 - uint64 in original \n\nName: proto \nDescription: The transport layer protocol of the connection. \nData type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset \n\nName: service \nDescription: An identification of an application protocol being sent over the connection. \nData type: optional string \n\nName: duration \nDescription: How long the connection lasted. \nData type: optional float64 - time interval \n\nName: orig_bytes \nDescription: The number of payload bytes the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_bytes \nDescription:The number of payload bytes the responder sent. \nData type: optional int64 - uint64 in original \n\nName: conn_state \nDescription: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH) \nData type: optional string \n\nName: missed_bytes \nDescription: Indicates the number of bytes missed in content gaps, which is representative of packet loss. \nData type: optional int64 - uint64 in original. default = 0\n\nName: history \nDescription: Records the state history of connections as a string of letters. \nData type: optional string \n\nName: orig_pkts \nDescription: Number of packets that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: orig_ip_bytes \nDescription: Number of IP level bytes that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_pkts \nDescription: Number of packets that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: resp_ip_bytes \nDescription: Number of IP level bytes that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: label \nDescription: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types \nData type: string - enum(Malicious, Benign)\n\nNOTE: ts, uid, id.orig_h, id.resp_h have been removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h) using this dataset, as that can lead to over fitting to dataset specific times and addresses. \nFurther local_orig, local_resp have been removed as they are null in all rows, so they are useless for training.\n</details>\n\nIf you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL”"
] | [
36,
303,
781
] | [
"passage: TAGS\n#task_categories-question-answering #task_categories-tabular-classification #language-English #code #region-us \n# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.\n\nThis dataset only notes if the data is Malcious or Benign. The original dataset labels the type of malcious traffic aswell. This means this processing of the dataset is only suited for binary classification."
] |
7b0927c307a94d3585de25821d89f44cd18b40d5 | # Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic
**Homepage:** [https://www.stratosphereips.org/datasets-iot23](https://www.stratosphereips.org/datasets-iot23)
This dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices [(Comparative Analysis of IoT Botnet Datasets)](https://doi.org/10.53070/bbd.1173687).
The selection of the subset was determined by [Aqeel Ahmed on Kaggle](https://www.kaggle.com/datasets/engraqeel/iot23preprocesseddata) and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of conn.log.labelled files.
# Feature information:
All features originate from the [Zeek](https://docs.zeek.org/en/master/scripts/base/protocols/conn/main.zeek.html#type-Conn::Info) processing performed by the dataset creators. [See notes here for caviats for each column](https://docs.zeek.org/en/master/scripts/base/protocols/conn/main.zeek.html#type-Conn::Info).
<details>
<summary>Expand for feature names, descriptions, and datatypes</summary>
Name: ts
Desription: This is the time of the first packet.
Data Type: float64 - Timestamp
Name: uid
Description: A Zeek-defined unique identifier of the connection.
Data type: string
Name: id.orig_h
Description: The originator’s IP address.
Data type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6
Name: id.orig_p
Description: The originator’s port number.
Data type: int64 - uint64 in original
Name: id.resp_h
Description: The responder’s IP address.
Data type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6
Name: id.resp_p
Description: The responder’s port number.
Data type: int64 - uint64 in original
Name: proto
Description: The transport layer protocol of the connection.
Data type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset
Name: service
Description: An identification of an application protocol being sent over the connection.
Data type: optional string
Name: duration
Description: How long the connection lasted.
Data type: optional float64 - time interval
Name: orig_bytes
Description: The number of payload bytes the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_bytes
Description:The number of payload bytes the responder sent.
Data type: optional int64 - uint64 in original
Name: conn_state
Description: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH)
Data type: optional string
Name: local_orig
Description: If the connection is originated locally, this value will be T. If it was originated remotely it will be F.
Data type: optional float64 - bool in original but null for all columns
Name: local_resp
Description: If the connection is responded to locally, this value will be T. If it was responded to remotely it will be F.
Data type: optional float64 - bool in original but null for all columns
Name: missed_bytes
Description: Indicates the number of bytes missed in content gaps, which is representative of packet loss.
Data type: optional int64 - uint64 in original. default = 0
Name: history
Description: Records the state history of connections as a string of letters.
Data type: optional string
Name: orig_pkts
Description: Number of packets that the originator sent.
Data type: optional int64 - uint64 in original
Name: orig_ip_bytes
Description: Number of IP level bytes that the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_pkts
Description: Number of packets that the responder sent.
Data type: optional int64 - uint64 in original
Name: resp_ip_bytes
Description: Number of IP level bytes that the responder sent.
Data type: optional int64 - uint64 in original
Name: label
Description: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types
Data type: string - enum('PartOfAHorizontalPortScan', 'Okiru', 'DDoS', 'C&C-HeartBeat',
'Benign', 'C&C-Torii', 'C&C', 'C&C-FileDownload', 'Okiru-Attack',
'Attack', 'FileDownload', 'C&C-HeartBeat-FileDownload',
'C&C-Mirai')
NOTE: ts, uid, id.orig_h, id.resp_h SHOULD BE removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h), as that can lead to over fitting to dataset specific times and addresses.
Further local_orig, local_resp SHOULD BE removed as they are null in all rows, so they are useless for training.
</details>
## Citation
If you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. http://doi.org/10.5281/zenodo.4743746”
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | yashika0998/iot-23-preprocessed-allcolumns | [
"task_categories:tabular-classification",
"task_categories:table-question-answering",
"language:en",
"code",
"region:us"
] | 2023-12-01T18:19:01+00:00 | {"language": ["en"], "task_categories": ["tabular-classification", "table-question-answering"], "dataset_info": {"features": [{"name": "ts", "dtype": "float64"}, {"name": "uid", "dtype": "string"}, {"name": "id.orig_h", "dtype": "string"}, {"name": "id.orig_p", "dtype": "int64"}, {"name": "id.resp_h", "dtype": "string"}, {"name": "id.resp_p", "dtype": "int64"}, {"name": "proto", "dtype": "string"}, {"name": "service", "dtype": "string"}, {"name": "duration", "dtype": "float64"}, {"name": "orig_bytes", "dtype": "int64"}, {"name": "resp_bytes", "dtype": "int64"}, {"name": "conn_state", "dtype": "string"}, {"name": "local_orig", "dtype": "float64"}, {"name": "local_resp", "dtype": "float64"}, {"name": "missed_bytes", "dtype": "int64"}, {"name": "history", "dtype": "string"}, {"name": "orig_pkts", "dtype": "int64"}, {"name": "orig_ip_bytes", "dtype": "int64"}, {"name": "resp_pkts", "dtype": "int64"}, {"name": "resp_ip_bytes", "dtype": "int64"}, {"name": "label", "dtype": "string"}], "splits": [{"name": "train", "num_bytes": 1232978140, "num_examples": 6046623}], "download_size": 274218995, "dataset_size": 1232978140}, "configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}]}], "tags": ["code"]} | 2023-12-01T18:32:45+00:00 | [] | [
"en"
] | TAGS
#task_categories-tabular-classification #task_categories-table-question-answering #language-English #code #region-us
| # Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic
Homepage: URL
This dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets).
The selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.
# Feature information:
All features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column.
<details>
<summary>Expand for feature names, descriptions, and datatypes</summary>
Name: ts
Desription: This is the time of the first packet.
Data Type: float64 - Timestamp
Name: uid
Description: A Zeek-defined unique identifier of the connection.
Data type: string
Name: id.orig_h
Description: The originator’s IP address.
Data type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6
Name: id.orig_p
Description: The originator’s port number.
Data type: int64 - uint64 in original
Name: id.resp_h
Description: The responder’s IP address.
Data type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6
Name: id.resp_p
Description: The responder’s port number.
Data type: int64 - uint64 in original
Name: proto
Description: The transport layer protocol of the connection.
Data type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset
Name: service
Description: An identification of an application protocol being sent over the connection.
Data type: optional string
Name: duration
Description: How long the connection lasted.
Data type: optional float64 - time interval
Name: orig_bytes
Description: The number of payload bytes the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_bytes
Description:The number of payload bytes the responder sent.
Data type: optional int64 - uint64 in original
Name: conn_state
Description: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH)
Data type: optional string
Name: local_orig
Description: If the connection is originated locally, this value will be T. If it was originated remotely it will be F.
Data type: optional float64 - bool in original but null for all columns
Name: local_resp
Description: If the connection is responded to locally, this value will be T. If it was responded to remotely it will be F.
Data type: optional float64 - bool in original but null for all columns
Name: missed_bytes
Description: Indicates the number of bytes missed in content gaps, which is representative of packet loss.
Data type: optional int64 - uint64 in original. default = 0
Name: history
Description: Records the state history of connections as a string of letters.
Data type: optional string
Name: orig_pkts
Description: Number of packets that the originator sent.
Data type: optional int64 - uint64 in original
Name: orig_ip_bytes
Description: Number of IP level bytes that the originator sent.
Data type: optional int64 - uint64 in original
Name: resp_pkts
Description: Number of packets that the responder sent.
Data type: optional int64 - uint64 in original
Name: resp_ip_bytes
Description: Number of IP level bytes that the responder sent.
Data type: optional int64 - uint64 in original
Name: label
Description: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types
Data type: string - enum('PartOfAHorizontalPortScan', 'Okiru', 'DDoS', 'C&C-HeartBeat',
'Benign', 'C&C-Torii', 'C&C', 'C&C-FileDownload', 'Okiru-Attack',
'Attack', 'FileDownload', 'C&C-HeartBeat-FileDownload',
'C&C-Mirai')
NOTE: ts, uid, id.orig_h, id.resp_h SHOULD BE removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h), as that can lead to over fitting to dataset specific times and addresses.
Further local_orig, local_resp SHOULD BE removed as they are null in all rows, so they are useless for training.
</details>
If you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL”
More Information needed | [
"# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.",
"# Feature information:\n\nAll features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column. \n<details>\n <summary>Expand for feature names, descriptions, and datatypes</summary>\n\nName: ts \nDesription: This is the time of the first packet. \nData Type: float64 - Timestamp \n\nName: uid \nDescription: A Zeek-defined unique identifier of the connection. \nData type: string \n\nName: id.orig_h \nDescription: The originator’s IP address. \nData type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6 \n\nName: id.orig_p \nDescription: The originator’s port number. \nData type: int64 - uint64 in original \n\nName: id.resp_h \nDescription: The responder’s IP address. \nData type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6 \n\nName: id.resp_p \nDescription: The responder’s port number. \nData type: int64 - uint64 in original \n\nName: proto \nDescription: The transport layer protocol of the connection. \nData type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset \n\nName: service \nDescription: An identification of an application protocol being sent over the connection. \nData type: optional string \n\nName: duration \nDescription: How long the connection lasted. \nData type: optional float64 - time interval \n\nName: orig_bytes \nDescription: The number of payload bytes the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_bytes \nDescription:The number of payload bytes the responder sent. \nData type: optional int64 - uint64 in original \n\nName: conn_state \nDescription: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH) \nData type: optional string \n\nName: local_orig \nDescription: If the connection is originated locally, this value will be T. If it was originated remotely it will be F.\nData type: optional float64 - bool in original but null for all columns\n\nName: local_resp\nDescription: If the connection is responded to locally, this value will be T. If it was responded to remotely it will be F.\nData type: optional float64 - bool in original but null for all columns\n\nName: missed_bytes \nDescription: Indicates the number of bytes missed in content gaps, which is representative of packet loss. \nData type: optional int64 - uint64 in original. default = 0\n\nName: history \nDescription: Records the state history of connections as a string of letters. \nData type: optional string \n\nName: orig_pkts \nDescription: Number of packets that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: orig_ip_bytes \nDescription: Number of IP level bytes that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_pkts \nDescription: Number of packets that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: resp_ip_bytes \nDescription: Number of IP level bytes that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: label \nDescription: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types \nData type: string - enum('PartOfAHorizontalPortScan', 'Okiru', 'DDoS', 'C&C-HeartBeat',\n 'Benign', 'C&C-Torii', 'C&C', 'C&C-FileDownload', 'Okiru-Attack',\n 'Attack', 'FileDownload', 'C&C-HeartBeat-FileDownload',\n 'C&C-Mirai')\n\nNOTE: ts, uid, id.orig_h, id.resp_h SHOULD BE removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h), as that can lead to over fitting to dataset specific times and addresses. \nFurther local_orig, local_resp SHOULD BE removed as they are null in all rows, so they are useless for training.\n</details>\n\nIf you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL”\nMore Information needed"
] | [
"TAGS\n#task_categories-tabular-classification #task_categories-table-question-answering #language-English #code #region-us \n",
"# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files.",
"# Feature information:\n\nAll features originate from the Zeek processing performed by the dataset creators. See notes here for caviats for each column. \n<details>\n <summary>Expand for feature names, descriptions, and datatypes</summary>\n\nName: ts \nDesription: This is the time of the first packet. \nData Type: float64 - Timestamp \n\nName: uid \nDescription: A Zeek-defined unique identifier of the connection. \nData type: string \n\nName: id.orig_h \nDescription: The originator’s IP address. \nData type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6 \n\nName: id.orig_p \nDescription: The originator’s port number. \nData type: int64 - uint64 in original \n\nName: id.resp_h \nDescription: The responder’s IP address. \nData type: string - for the form 255.255.255.255 for IPv4 or [aaaa:bbbb:cccc:dddd:eeee:ffff:1111:2222] for IPv6 \n\nName: id.resp_p \nDescription: The responder’s port number. \nData type: int64 - uint64 in original \n\nName: proto \nDescription: The transport layer protocol of the connection. \nData type: string - enum(unknown_transport, tcp, udp, icmp). Only TCP and UDP in subset \n\nName: service \nDescription: An identification of an application protocol being sent over the connection. \nData type: optional string \n\nName: duration \nDescription: How long the connection lasted. \nData type: optional float64 - time interval \n\nName: orig_bytes \nDescription: The number of payload bytes the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_bytes \nDescription:The number of payload bytes the responder sent. \nData type: optional int64 - uint64 in original \n\nName: conn_state \nDescription: Value indicating connection state. (S0, S1, SF, REJ, S2, S3, RSTO, RSTR, RSTOS0, RSTRH, SH, SHR, OTH) \nData type: optional string \n\nName: local_orig \nDescription: If the connection is originated locally, this value will be T. If it was originated remotely it will be F.\nData type: optional float64 - bool in original but null for all columns\n\nName: local_resp\nDescription: If the connection is responded to locally, this value will be T. If it was responded to remotely it will be F.\nData type: optional float64 - bool in original but null for all columns\n\nName: missed_bytes \nDescription: Indicates the number of bytes missed in content gaps, which is representative of packet loss. \nData type: optional int64 - uint64 in original. default = 0\n\nName: history \nDescription: Records the state history of connections as a string of letters. \nData type: optional string \n\nName: orig_pkts \nDescription: Number of packets that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: orig_ip_bytes \nDescription: Number of IP level bytes that the originator sent. \nData type: optional int64 - uint64 in original \n\nName: resp_pkts \nDescription: Number of packets that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: resp_ip_bytes \nDescription: Number of IP level bytes that the responder sent. \nData type: optional int64 - uint64 in original \n\nName: label \nDescription: Specifies if data point is benign or some form of malicious. See the dataset creators paper for descriptions of attack types \nData type: string - enum('PartOfAHorizontalPortScan', 'Okiru', 'DDoS', 'C&C-HeartBeat',\n 'Benign', 'C&C-Torii', 'C&C', 'C&C-FileDownload', 'Okiru-Attack',\n 'Attack', 'FileDownload', 'C&C-HeartBeat-FileDownload',\n 'C&C-Mirai')\n\nNOTE: ts, uid, id.orig_h, id.resp_h SHOULD BE removed as they are dataset specific. Models should not be trained with specific timestamps or IP addresses (id.orig_h), as that can lead to over fitting to dataset specific times and addresses. \nFurther local_orig, local_resp SHOULD BE removed as they are null in all rows, so they are useless for training.\n</details>\n\nIf you are using this dataset for your research, please reference it as “Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga. (2020). IoT-23: A labeled dataset with malicious and benign IoT network traffic (Version 1.0.0) [Data set]. Zenodo. URL”\nMore Information needed"
] | [
38,
253,
1171
] | [
"passage: TAGS\n#task_categories-tabular-classification #task_categories-table-question-answering #language-English #code #region-us \n# Aposemat IoT-23 - a Labeled Dataset with Malcious and Benign Iot Network Traffic \nHomepage: URL\n\nThis dataset contains a subset of the data from 20 captures of Malcious network traffic and 3 captures from live Benign Traffic on Internet of Things (IoT) devices. Created by Sebastian Garcia, Agustin Parmisano, & Maria Jose Erquiaga at the Avast AIC laboratory with the funding of Avast Software, this dataset is one of the best in the field for Intrusion Detection Systems (IDS) for IoT Devices (Comparative Analysis of IoT Botnet Datasets). \n\nThe selection of the subset was determined by Aqeel Ahmed on Kaggle and contains 6 million samples. The Kaggle upload, nor this one, have employed data balancing. The Kaggle card does not contain methodology to understand what criteria was used to select these samples. If you want ensure best practice, use this dataset to mock-up processing the data into a model before using the full dataset with data balancing. This will require processing the 8GB of URL.labelled files."
] |
56322d544010ccda54f68c2f8e3beb14203a1efa | # Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
This dataset card aims to be a base template for new datasets. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
## Dataset Details
### Dataset Description
<!-- Provide a longer summary of what this dataset is. -->
- **Curated by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
### Dataset Sources [optional]
<!-- Provide the basic links for the dataset. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the dataset is intended to be used. -->
### Direct Use
<!-- This section describes suitable use cases for the dataset. -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
[More Information Needed]
## Dataset Structure
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
[More Information Needed]
## Dataset Creation
### Curation Rationale
<!-- Motivation for the creation of this dataset. -->
[More Information Needed]
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
#### Data Collection and Processing
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
[More Information Needed]
#### Who are the source data producers?
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
[More Information Needed]
### Annotations [optional]
<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
#### Annotation process
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
[More Information Needed]
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
[More Information Needed]
#### Personal and Sensitive Information
<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Dataset Card Authors [optional]
[More Information Needed]
## Dataset Card Contact
[More Information Needed] | Dnsibu/serials | [
"task_categories:token-classification",
"size_categories:n<1K",
"language:en",
"region:us"
] | 2023-12-01T19:27:40+00:00 | {"language": ["en"], "size_categories": ["n<1K"], "task_categories": ["token-classification"], "pretty_name": "serials"} | 2023-12-12T21:46:06+00:00 | [] | [
"en"
] | TAGS
#task_categories-token-classification #size_categories-n<1K #language-English #region-us
| # Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
## Dataset Details
### Dataset Description
- Curated by:
- Funded by [optional]:
- Shared by [optional]:
- Language(s) (NLP):
- License:
### Dataset Sources [optional]
- Repository:
- Paper [optional]:
- Demo [optional]:
## Uses
### Direct Use
### Out-of-Scope Use
## Dataset Structure
## Dataset Creation
### Curation Rationale
### Source Data
#### Data Collection and Processing
#### Who are the source data producers?
### Annotations [optional]
#### Annotation process
#### Who are the annotators?
#### Personal and Sensitive Information
## Bias, Risks, and Limitations
### Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
[optional]
BibTeX:
APA:
## Glossary [optional]
## More Information [optional]
## Dataset Card Authors [optional]
## Dataset Card Contact
| [
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
"TAGS\n#task_categories-token-classification #size_categories-n<1K #language-English #region-us \n",
"# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.",
"## Dataset Details",
"### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:",
"### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:",
"## Uses",
"### Direct Use",
"### Out-of-Scope Use",
"## Dataset Structure",
"## Dataset Creation",
"### Curation Rationale",
"### Source Data",
"#### Data Collection and Processing",
"#### Who are the source data producers?",
"### Annotations [optional]",
"#### Annotation process",
"#### Who are the annotators?",
"#### Personal and Sensitive Information",
"## Bias, Risks, and Limitations",
"### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:",
"## Glossary [optional]",
"## More Information [optional]",
"## Dataset Card Authors [optional]",
"## Dataset Card Contact"
] | [
32,
34,
4,
40,
29,
3,
4,
9,
6,
5,
7,
4,
7,
10,
9,
5,
9,
8,
10,
46,
8,
7,
10,
5
] | [
"passage: TAGS\n#task_categories-token-classification #size_categories-n<1K #language-English #region-us \n# Dataset Card for Dataset Name\n\n\n\nThis dataset card aims to be a base template for new datasets. It has been generated using this raw template.## Dataset Details### Dataset Description\n\n\n\n\n\n- Curated by: \n- Funded by [optional]: \n- Shared by [optional]: \n- Language(s) (NLP): \n- License:### Dataset Sources [optional]\n\n\n\n- Repository: \n- Paper [optional]: \n- Demo [optional]:## Uses### Direct Use### Out-of-Scope Use## Dataset Structure## Dataset Creation### Curation Rationale### Source Data#### Data Collection and Processing#### Who are the source data producers?### Annotations [optional]#### Annotation process#### Who are the annotators?#### Personal and Sensitive Information## Bias, Risks, and Limitations### Recommendations\n\n\n\nUsers should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.\n\n[optional]\n\n\n\nBibTeX:\n\n\n\nAPA:## Glossary [optional]## More Information [optional]## Dataset Card Authors [optional]## Dataset Card Contact"
] |
b4ac8bab3bb80222a3ed7332589112d7e1e64280 |
### Dataset Description
This media storms dataset contains metadata and full text descriptions of all news articles identified as media storms in our paper "When it Rains it Pours: Modeling Media Storms and the News Ecosystem".
For only the news articles identified in media storms, use the smaller mediaStormArticles.tsv file. For all articles in NELA-GT-2020, NELA-GT-2021, and NELA-Local with story-cluster labels, use the larger storyClusterArticles.tsv.gz file.
- **Curated by:** Ben Litterer, David Jurgens, Dallas Card. Original data curated by creators of NELA-GT-2020, NELA-GT-2021, and NELA-Local (associated papers linked below).
## Uses
This data can be used to model media storms (large, ubiquitous, long lasting stories) between April 2020 - December 2021.
## Dataset Structure
For dataset field descriptions, reference documentation for `NELA-GT-2020`, `NELA-GT-2021`, and `NELA-local`. Full text is contained in `content` and labels indicating which media storm
an article took part in are indicated in `stormID`.
### Source Data
The underlying data for this dataset are from `NELA-GT-2020`, `NELA-GT-2021`, and `NELA-local`. We have identified media storms in these data and added cluster labels indicating
which articles took part in which media storms.
- NELA-GT-2020: https://arxiv.org/abs/2102.04567
- NELA-GT-2021: https://arxiv.org/abs/2203.05659
- NELA-Local: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwigm4qdgO-CAxUYkIkEHedwCWcQFnoECA0QAQ&url=https%3A%2F%2Fojs.aaai.org%2Findex.php%2FICWSM%2Farticle%2Fview%2F19379&usg=AOvVaw3BqIHBBqwP-vFutUywo9KW&opi=89978449 | Blablablab/mediaStorms | [
"arxiv:2102.04567",
"arxiv:2203.05659",
"region:us"
] | 2023-12-01T19:36:30+00:00 | {} | 2023-12-26T20:44:11+00:00 | [
"2102.04567",
"2203.05659"
] | [] | TAGS
#arxiv-2102.04567 #arxiv-2203.05659 #region-us
|
### Dataset Description
This media storms dataset contains metadata and full text descriptions of all news articles identified as media storms in our paper "When it Rains it Pours: Modeling Media Storms and the News Ecosystem".
For only the news articles identified in media storms, use the smaller URL file. For all articles in NELA-GT-2020, NELA-GT-2021, and NELA-Local with story-cluster labels, use the larger URL file.
- Curated by: Ben Litterer, David Jurgens, Dallas Card. Original data curated by creators of NELA-GT-2020, NELA-GT-2021, and NELA-Local (associated papers linked below).
## Uses
This data can be used to model media storms (large, ubiquitous, long lasting stories) between April 2020 - December 2021.
## Dataset Structure
For dataset field descriptions, reference documentation for 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. Full text is contained in 'content' and labels indicating which media storm
an article took part in are indicated in 'stormID'.
### Source Data
The underlying data for this dataset are from 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. We have identified media storms in these data and added cluster labels indicating
which articles took part in which media storms.
- NELA-GT-2020: URL
- NELA-GT-2021: URL
- NELA-Local: URL | [
"### Dataset Description\n\nThis media storms dataset contains metadata and full text descriptions of all news articles identified as media storms in our paper \"When it Rains it Pours: Modeling Media Storms and the News Ecosystem\". \n\nFor only the news articles identified in media storms, use the smaller URL file. For all articles in NELA-GT-2020, NELA-GT-2021, and NELA-Local with story-cluster labels, use the larger URL file. \n\n- Curated by: Ben Litterer, David Jurgens, Dallas Card. Original data curated by creators of NELA-GT-2020, NELA-GT-2021, and NELA-Local (associated papers linked below).",
"## Uses\n\nThis data can be used to model media storms (large, ubiquitous, long lasting stories) between April 2020 - December 2021.",
"## Dataset Structure\n\nFor dataset field descriptions, reference documentation for 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. Full text is contained in 'content' and labels indicating which media storm \nan article took part in are indicated in 'stormID'.",
"### Source Data\n\nThe underlying data for this dataset are from 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. We have identified media storms in these data and added cluster labels indicating \nwhich articles took part in which media storms. \n\n- NELA-GT-2020: URL\n- NELA-GT-2021: URL\n- NELA-Local: URL"
] | [
"TAGS\n#arxiv-2102.04567 #arxiv-2203.05659 #region-us \n",
"### Dataset Description\n\nThis media storms dataset contains metadata and full text descriptions of all news articles identified as media storms in our paper \"When it Rains it Pours: Modeling Media Storms and the News Ecosystem\". \n\nFor only the news articles identified in media storms, use the smaller URL file. For all articles in NELA-GT-2020, NELA-GT-2021, and NELA-Local with story-cluster labels, use the larger URL file. \n\n- Curated by: Ben Litterer, David Jurgens, Dallas Card. Original data curated by creators of NELA-GT-2020, NELA-GT-2021, and NELA-Local (associated papers linked below).",
"## Uses\n\nThis data can be used to model media storms (large, ubiquitous, long lasting stories) between April 2020 - December 2021.",
"## Dataset Structure\n\nFor dataset field descriptions, reference documentation for 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. Full text is contained in 'content' and labels indicating which media storm \nan article took part in are indicated in 'stormID'.",
"### Source Data\n\nThe underlying data for this dataset are from 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. We have identified media storms in these data and added cluster labels indicating \nwhich articles took part in which media storms. \n\n- NELA-GT-2020: URL\n- NELA-GT-2021: URL\n- NELA-Local: URL"
] | [
22,
157,
34,
72,
90
] | [
"passage: TAGS\n#arxiv-2102.04567 #arxiv-2203.05659 #region-us \n### Dataset Description\n\nThis media storms dataset contains metadata and full text descriptions of all news articles identified as media storms in our paper \"When it Rains it Pours: Modeling Media Storms and the News Ecosystem\". \n\nFor only the news articles identified in media storms, use the smaller URL file. For all articles in NELA-GT-2020, NELA-GT-2021, and NELA-Local with story-cluster labels, use the larger URL file. \n\n- Curated by: Ben Litterer, David Jurgens, Dallas Card. Original data curated by creators of NELA-GT-2020, NELA-GT-2021, and NELA-Local (associated papers linked below).## Uses\n\nThis data can be used to model media storms (large, ubiquitous, long lasting stories) between April 2020 - December 2021.## Dataset Structure\n\nFor dataset field descriptions, reference documentation for 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. Full text is contained in 'content' and labels indicating which media storm \nan article took part in are indicated in 'stormID'.### Source Data\n\nThe underlying data for this dataset are from 'NELA-GT-2020', 'NELA-GT-2021', and 'NELA-local'. We have identified media storms in these data and added cluster labels indicating \nwhich articles took part in which media storms. \n\n- NELA-GT-2020: URL\n- NELA-GT-2021: URL\n- NELA-Local: URL"
] |
a28cdbff6f8f18806f13ee88e4897d85d2ce0678 | # Dataset Card for "tokenized_dataset_bart"
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards) | witchling22/tokenized_dataset_bart | [
"region:us"
] | 2023-12-01T20:10:10+00:00 | {"configs": [{"config_name": "default", "data_files": [{"split": "train", "path": "data/train-*"}, {"split": "test", "path": "data/test-*"}, {"split": "validation", "path": "data/validation-*"}]}], "dataset_info": {"features": [{"name": "source", "dtype": "string"}, {"name": "source_labels", "dtype": "string"}, {"name": "rouge_scores", "dtype": "string"}, {"name": "paper_id", "dtype": "string"}, {"name": "target", "dtype": "string"}, {"name": "full_source_text", "dtype": "string"}, {"name": "input_ids", "sequence": "int32"}, {"name": "attention_mask", "sequence": "int8"}, {"name": "labels", "sequence": "int64"}], "splits": [{"name": "train", "num_bytes": 17340567, "num_examples": 1992}, {"name": "test", "num_bytes": 5620222, "num_examples": 618}, {"name": "validation", "num_bytes": 5534448, "num_examples": 619}], "download_size": 6328102, "dataset_size": 28495237}} | 2023-12-01T20:10:14+00:00 | [] | [] | TAGS
#region-us
| # Dataset Card for "tokenized_dataset_bart"
More Information needed | [
"# Dataset Card for \"tokenized_dataset_bart\"\n\nMore Information needed"
] | [
"TAGS\n#region-us \n",
"# Dataset Card for \"tokenized_dataset_bart\"\n\nMore Information needed"
] | [
6,
18
] | [
"passage: TAGS\n#region-us \n# Dataset Card for \"tokenized_dataset_bart\"\n\nMore Information needed"
] |
4cdd7e80b4a74048e403687cd01d4807c530c223 |
This is a data on fake news and for elections 2024 , specially curated to highlight racial slurs and discriminations in political speeches.
https://arxiv.org/abs/2312.03750
## Citation
If you use this dataset in your research, please cite it as follows:
```bibtex
@article{rahman2023analyzing,
title={Analyzing the Impact of Fake News on the Anticipated Outcome of the 2024 Election Ahead of Time},
author={Raza, Shaina and Rahman, Mizanur and Ghuge, Shardul},
journal={arXiv preprint arXiv:2312.03750},
year={2023}
}
| newsmediabias/fake_news_elections2024 | [
"license:openrail",
"arxiv:2312.03750",
"region:us"
] | 2023-12-01T20:21:28+00:00 | {"license": "openrail"} | 2024-01-06T17:31:22+00:00 | [
"2312.03750"
] | [] | TAGS
#license-openrail #arxiv-2312.03750 #region-us
|
This is a data on fake news and for elections 2024 , specially curated to highlight racial slurs and discriminations in political speeches.
URL
If you use this dataset in your research, please cite it as follows:
'''bibtex
@article{rahman2023analyzing,
title={Analyzing the Impact of Fake News on the Anticipated Outcome of the 2024 Election Ahead of Time},
author={Raza, Shaina and Rahman, Mizanur and Ghuge, Shardul},
journal={arXiv preprint arXiv:2312.03750},
year={2023}
}
| [] | [
"TAGS\n#license-openrail #arxiv-2312.03750 #region-us \n"
] | [
21
] | [
"passage: TAGS\n#license-openrail #arxiv-2312.03750 #region-us \n"
] |
c652513690a97f317a6000f85a5111b71686c798 |
# SOCRATIS: A benchmark of diverse open-ended emotional reactions to image-caption pairs.
### ICCV WECIA Workshop 2023 (oral)
[Project Page](https://kdeng55.github.io/socratis-website/), [Paper](https://arxiv.org/abs/2308.16741)
We release a benchmark which contains 18K diverse emotions and reasons for feeling them on 2K image-caption pairs.
Our current preliminary findings have shown that Humans prefer human-written emotional reactions over machine-generated by more than two times.
We also find that current metrics fail to correlate with human preference, leaving significant room for research!
We release the data publicly.
`test.json` contains the testing data in the following format:
```
{
unique_id: [[image_path, caption, emotions, explanations, anonymized_demographics], ...]
}
```
The `unique_id` is a unique id for a image-caption pair. Each `unique_id` key has a list of entries from diverse workers.
Each entry consists of the emotions and the explanations for feeling that emotion. Demographics may be missing for many annotations since they were optional and some workers opted to not disclose it. All data is anonymized.
The image files are at: https://drive.google.com/file/d/1J8SiUEfKqc5rfxE1nwZUrG1Hcz7Djc3G/view?usp=sharing. | array/socratis_image_text_emotion | [
"task_categories:text-classification",
"task_categories:image-classification",
"task_categories:conversational",
"license:mit",
"arxiv:2308.16741",
"region:us"
] | 2023-12-01T20:25:13+00:00 | {"license": "mit", "task_categories": ["text-classification", "image-classification", "conversational"]} | 2023-12-01T20:47:36+00:00 | [
"2308.16741"
] | [] | TAGS
#task_categories-text-classification #task_categories-image-classification #task_categories-conversational #license-mit #arxiv-2308.16741 #region-us
|
# SOCRATIS: A benchmark of diverse open-ended emotional reactions to image-caption pairs.
### ICCV WECIA Workshop 2023 (oral)
Project Page, Paper
We release a benchmark which contains 18K diverse emotions and reasons for feeling them on 2K image-caption pairs.
Our current preliminary findings have shown that Humans prefer human-written emotional reactions over machine-generated by more than two times.
We also find that current metrics fail to correlate with human preference, leaving significant room for research!
We release the data publicly.
'URL' contains the testing data in the following format:
The 'unique_id' is a unique id for a image-caption pair. Each 'unique_id' key has a list of entries from diverse workers.
Each entry consists of the emotions and the explanations for feeling that emotion. Demographics may be missing for many annotations since they were optional and some workers opted to not disclose it. All data is anonymized.
The image files are at: URL | [
"# SOCRATIS: A benchmark of diverse open-ended emotional reactions to image-caption pairs.",
"### ICCV WECIA Workshop 2023 (oral)\n\nProject Page, Paper \n\nWe release a benchmark which contains 18K diverse emotions and reasons for feeling them on 2K image-caption pairs. \n\nOur current preliminary findings have shown that Humans prefer human-written emotional reactions over machine-generated by more than two times.\nWe also find that current metrics fail to correlate with human preference, leaving significant room for research!\n\nWe release the data publicly.\n\n'URL' contains the testing data in the following format:\n\n\n\nThe 'unique_id' is a unique id for a image-caption pair. Each 'unique_id' key has a list of entries from diverse workers.\n\nEach entry consists of the emotions and the explanations for feeling that emotion. Demographics may be missing for many annotations since they were optional and some workers opted to not disclose it. All data is anonymized. \n\nThe image files are at: URL"
] | [
"TAGS\n#task_categories-text-classification #task_categories-image-classification #task_categories-conversational #license-mit #arxiv-2308.16741 #region-us \n",
"# SOCRATIS: A benchmark of diverse open-ended emotional reactions to image-caption pairs.",
"### ICCV WECIA Workshop 2023 (oral)\n\nProject Page, Paper \n\nWe release a benchmark which contains 18K diverse emotions and reasons for feeling them on 2K image-caption pairs. \n\nOur current preliminary findings have shown that Humans prefer human-written emotional reactions over machine-generated by more than two times.\nWe also find that current metrics fail to correlate with human preference, leaving significant room for research!\n\nWe release the data publicly.\n\n'URL' contains the testing data in the following format:\n\n\n\nThe 'unique_id' is a unique id for a image-caption pair. Each 'unique_id' key has a list of entries from diverse workers.\n\nEach entry consists of the emotions and the explanations for feeling that emotion. Demographics may be missing for many annotations since they were optional and some workers opted to not disclose it. All data is anonymized. \n\nThe image files are at: URL"
] | [
51,
24,
207
] | [
"passage: TAGS\n#task_categories-text-classification #task_categories-image-classification #task_categories-conversational #license-mit #arxiv-2308.16741 #region-us \n# SOCRATIS: A benchmark of diverse open-ended emotional reactions to image-caption pairs.### ICCV WECIA Workshop 2023 (oral)\n\nProject Page, Paper \n\nWe release a benchmark which contains 18K diverse emotions and reasons for feeling them on 2K image-caption pairs. \n\nOur current preliminary findings have shown that Humans prefer human-written emotional reactions over machine-generated by more than two times.\nWe also find that current metrics fail to correlate with human preference, leaving significant room for research!\n\nWe release the data publicly.\n\n'URL' contains the testing data in the following format:\n\n\n\nThe 'unique_id' is a unique id for a image-caption pair. Each 'unique_id' key has a list of entries from diverse workers.\n\nEach entry consists of the emotions and the explanations for feeling that emotion. Demographics may be missing for many annotations since they were optional and some workers opted to not disclose it. All data is anonymized. \n\nThe image files are at: URL"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.