Hugging Face's logo

Datasets:
2A2I
/
Arabic_Aya

Tasks:
Text Classification
Translation
Summarization
Modalities:
Tabular
Text
Languages:
Arabic
Size:
10M - 100M
ArXiv:
License:
Dataset card Viewer Files Community
3
Ali Elfilali commited on
Commit
7ecd35f
•
1 Parent(s): 3b8f6db

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +290 -118
README.md CHANGED
@@ -842,138 +842,310 @@ pretty_name: 2A
842
  size_categories:
843
  - 1M<n<10M
844
  ---
845
- # Dataset Card for Dataset Name
846
 
847
  <!-- Provide a quick summary of the dataset. -->
848
 
849
- 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).
850
 
851
- ## Dataset Details
852
 
853
- ### Dataset Description
854
 
855
- <!-- Provide a longer summary of what this dataset is. -->
 
856
 
857
- - **Curated by:** [More Information Needed]
858
- - **Funded by [optional]:** [More Information Needed]
859
- - **Shared by [optional]:** [More Information Needed]
860
- - **Language(s) (NLP):** [More Information Needed]
861
- - **License:** [More Information Needed]
862
 
863
- ### Dataset Sources [optional]
 
 
 
 
 
 
864
 
865
- <!-- Provide the basic links for the dataset. -->
 
866
 
867
- - **Repository:** [More Information Needed]
868
- - **Paper [optional]:** [More Information Needed]
869
- - **Demo [optional]:** [More Information Needed]
870
 
871
- ## Uses
 
872
 
873
- <!-- Address questions around how the dataset is intended to be used. -->
 
 
874
 
875
- ### Direct Use
876
 
877
- <!-- This section describes suitable use cases for the dataset. -->
 
878
 
879
- [More Information Needed]
 
880
 
881
- ### Out-of-Scope Use
882
 
883
- <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
884
-
885
- [More Information Needed]
886
-
887
- ## Dataset Structure
888
-
889
- <!-- 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. -->
890
-
891
- [More Information Needed]
892
-
893
- ## Dataset Creation
894
-
895
- ### Curation Rationale
896
-
897
- <!-- Motivation for the creation of this dataset. -->
898
-
899
- [More Information Needed]
900
-
901
- ### Source Data
902
-
903
- <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
904
-
905
- #### Data Collection and Processing
906
-
907
- <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
908
-
909
- [More Information Needed]
910
-
911
- #### Who are the source data producers?
912
-
913
- <!-- 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. -->
914
-
915
- [More Information Needed]
916
-
917
- ### Annotations [optional]
918
-
919
- <!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
920
-
921
- #### Annotation process
922
-
923
- <!-- 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. -->
924
-
925
- [More Information Needed]
926
-
927
- #### Who are the annotators?
928
-
929
- <!-- This section describes the people or systems who created the annotations. -->
930
-
931
- [More Information Needed]
932
-
933
- #### Personal and Sensitive Information
934
-
935
- <!-- 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. -->
936
-
937
- [More Information Needed]
938
-
939
- ## Bias, Risks, and Limitations
940
-
941
- <!-- This section is meant to convey both technical and sociotechnical limitations. -->
942
-
943
- [More Information Needed]
944
-
945
- ### Recommendations
946
-
947
- <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
948
-
949
- Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
950
-
951
- ## Citation [optional]
952
-
953
- <!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
954
-
955
- **BibTeX:**
956
-
957
- [More Information Needed]
958
-
959
- **APA:**
960
-
961
- [More Information Needed]
962
-
963
- ## Glossary [optional]
964
-
965
- <!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
966
-
967
- [More Information Needed]
968
-
969
- ## More Information [optional]
970
-
971
- [More Information Needed]
972
-
973
- ## Dataset Card Authors [optional]
974
-
975
- [More Information Needed]
976
-
977
- ## Dataset Card Contact
978
-
979
- [More Information Needed]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
842
  size_categories:
843
  - 1M<n<10M
844
  ---
845
+ # Dataset Card for : Arabic Aya (2A)
846
 
847
  <!-- Provide a quick summary of the dataset. -->
848
 
849
+ <!-- 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).-->
850
 
851
+ ## Dataset Description
852
 
853
+ **Arabic Aya (2A) : A Curated Subset of the Aya Collection for Arabic Language Processing**
854
 
855
+ ### Overview
856
+ `Arabic Aya` is a meticulously curated dataset derived from the comprehensive Aya collection by [CohereForAI](https://huggingface.co/CohereForAI), specifically focusing on Arabic text data. This dataset aggregates content from the [CohereForAI/aya_collection](https://huggingface.co/datasets/CohereForAI/aya_collection), [CohereForAI/aya_dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset), and [CohereForAI/aya_evaluation_suite](https://huggingface.co/datasets/CohereForAI/aya_evaluation_suite), filtering out all but the Arabic content, including both Modern Standard Arabic (MSA) and various regional dialects.
857
 
858
+ ### Purpose
859
+ The aim of 'Arabic Aya' is to provide researchers, technologists, and linguists with a ready-to-use Arabic text resource, significantly reducing the time and effort required for data preprocessing in NLP and AI projects focused on the Arabic language.
860
+ - Use the Aya datasets out of the box for your Arabic applications and research 😀
 
 
861
 
862
+ ### Dataset Sources & Infos
863
+ - **Data Origin**: Derived from 69 subsets of the original Aya datasets : [CohereForAI/aya_collection](https://huggingface.co/datasets/CohereForAI/aya_collection), [CohereForAI/aya_dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset), and [CohereForAI/aya_evaluation_suite](https://huggingface.co/datasets/CohereForAI/aya_evaluation_suite).
864
+ - **Languages**: Modern Standard Arabic (MSA) and a variety of Arabic dialects ( 'arb', 'arz', 'ary', 'ars', 'knc', 'acm', 'apc', 'aeb', 'ajp', 'acq' )
865
+ - **Applications**: Ideal for tasks such as language modeling, text classification, sentiment analysis, dialect identification, and machine translation.
866
+ - **Paper:** [2402.06619](https://huggingface.co/papers/2402.06619)
867
+ - **Maintainer:** [Elfilali Ali](https://huggingface.co/Ali-C137)
868
+ - **License:** Apache-2.0
869
 
870
+ ### Usage
871
+ This dataset serves as a foundational tool for those embarking on Arabic language projects, from academic research to commercial applications. By providing a pre-filtered source of Arabic text, 'Arabic Aya' enables users to dive straight into model training, analysis, and application development without the preliminary hassle of data cleaning and language filtering.
872
 
873
+ #### Use with HF Datasets library
874
+ To load this dataset with Datasets, you'll need to install Datasets as `pip install datasets --upgrade` and then use a similar code to the following:
 
875
 
876
+ ```python
877
+ from datasets import load_dataset
878
 
879
+ dataset = load_dataset("2A2I/Arabic_Aya", "CohereForAI-aya_collection-templated_mintaka")
880
+ ```
881
+ In the above code snippet, "CohereForAI-aya_collection-templated_mintaka" refers to the arabic version (100k rows) of the original "templated_mintaka" subset (780k rows) of the aya_collection. You can load other subsets by specifying its name at the time of loading the dataset.
882
 
 
883
 
884
+ ### Access and Contribution
885
+ Available on the Hugging Face Hub under [2A2I/Arabic_Aya](https://huggingface.co/datasets/2A2I/Arabic_Aya), 'Arabic Aya' invites contributions from the community. Users are encouraged to offer feedback, suggest improvements.
886
 
887
+ ### Support and Collaboration
888
+ We are committed to fostering an inclusive and supportive environment around Arabic AI and NLP research. For support, collaboration, or queries regarding the dataset, please reach out through the Hugging Face Hub's discussion section or reach out at [2A2I Contact Email](arabic.ai.initiative@gmail.com).
889
 
 
890
 
891
+
892
+
893
+ # Original Dataset Card of Aya by CohereForAI
894
+
895
+
896
+ ![Aya Header](https://huggingface.co/datasets/CohereForAI/aya_collection/resolve/main/aya_header.png)
897
+
898
+ # Dataset Summary
899
+ The Aya Collection is a massive multilingual collection consisting of 513 million instances of prompts and completions covering a wide range of tasks.
900
+ This collection incorporates instruction-style templates from fluent speakers and applies them to a curated list of datasets, as well as translations of instruction-style datasets into 101 languages. Aya Dataset, a human-curated multilingual instruction and response dataset, is also part of this collection. See our paper for more details regarding the collection.
901
+
902
+ - **Curated by:** Contributors of [Aya Open Science Intiative](https://cohere.com/research/aya)
903
+ - **Language(s):** 115 languages
904
+ - **License:** [Apache 2.0](https://opensource.org/license/apache-2-0)
905
+
906
+ - **Aya Datasets Family:**
907
+ | Name | Explanation |
908
+ |------|--------------|
909
+ | [aya_dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) | Human-annotated multilingual instruction finetuning dataset, comprising over 204K instances across 65 languages. |
910
+ | [aya_collection](https://huggingface.co/datasets/CohereForAI/aya_collection) | Created by applying instruction-style templates from fluent speakers to 44 datasets, including translations of 19 instruction-style datasets into 101 languages.|
911
+ | [aya_evaluation_suite](https://huggingface.co/datasets/CohereForAI/aya_evaluation_suite) | A diverse evaluation set for multilingual open-ended generation, featuring 250 culturally grounded prompts in 7 languages, 200 translated prompts in 24 languages, and human-edited versions selected for cross-cultural relevance from English Dolly in 6 languages.|
912
+
913
+
914
+ # Dataset
915
+ The `Aya Collection` is a comprehensive, large corpus of datasets that can be used by researchers around the world to train multilingual models. Our goal is only to include datasets with permissive licensing for manipulation and redistribution.
916
+
917
+ The `Aya Collection` consists of three different sources of data:
918
+
919
+ 1. Templated data: We collaborated with fluent speakers to create templates that allowed for the automatic expansion of existing datasets into various languages.
920
+ 2. Translated data: We translated a hand-selected subset of 19 datasets into 101 languages (114 dialects) using the NLLB 3.3B parameter machine translation model.
921
+ 3. Aya Dataset: We release the [Aya Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) as a subset of the overall collection. This is the only dataset in the collection that is human-annotated in its entirety.
922
+
923
+ ## Load with Datasets
924
+ To load this dataset with Datasets, you'll need to install Datasets as `pip install datasets --upgrade` and then use the following code:
925
+
926
+ ```python
927
+ from datasets import load_dataset
928
+
929
+ dataset = load_dataset("CohereForAI/aya_collection", "templated_mintaka")
930
+ ```
931
+ In the above code snippet, "templated_mintaka" refers to a subset of the aya_collection. You can load other subsets by specifying its name at the time of loading the dataset.
932
+
933
+ ## Data Instances
934
+ An example of a `train` instance looks as follows:
935
+ ```json
936
+ {'id': 246001,
937
+ 'inputs': 'The following query in English is taken from the geography category. What could be the answer to the question?\nWhat is the seventh tallest mountain in North America?',
938
+ 'targets': 'The answer is Mount Lucania.',
939
+ 'dataset_name': 'Mintaka-inst',
940
+ 'sub_dataset_name': '-',
941
+ 'task_type': 'question-answering',
942
+ 'template_id': 3,
943
+ 'language': 'eng',
944
+ 'split': 'train',
945
+ 'script': 'Latn'
946
+ }
947
+ ```
948
+
949
+ ## Data Fields
950
+ The data fields are the same among all splits:
951
+ - `id:` Unique id of the data point
952
+ - `inputs:` Prompt or input to the language model.
953
+ - `targets:` Completion or output of the language model.
954
+ - `dataset_name:` The name of the source dataset that the data point was taken from
955
+ - `sub_dataset_name:` If the source is a collection, this field indicates which part of that collection the data point was taken from. If it is not a collection, this field is left blank.
956
+ - `task_type:` The task type that this conversation belongs to.
957
+ - `template_id`: The id of the template applied to this data point.
958
+ - `language:` The ISO code of the dialect of the conversation.
959
+ - `script:` The script of the language.
960
+ - `split:` Indicates whether the data point is part of the `train` or the `test` split.
961
+
962
+
963
+
964
+ ### Statistics
965
+ The total number of data points, including the Aya Dataset` is 513,758,189. To view the breakdown of dialect codes and the respective templated and translated data point counts in the Aya Collection , refer to the toggled table below.
966
+
967
+ <details>
968
+ <summary> <b> Breakdown of Aya Collection data point counts grouped by dialects </b> </summary>
969
+
970
+ |dialect code|language|translated data point count|templated data point count|total count |
971
+ |------------|--------|---------------------------|--------------------------|---------------|
972
+ |ace |Achinese|8240684 |2000 |8242684 |
973
+ |acm |Arabic |4120342 |0 |4120342 |
974
+ |acq |Arabic |4120342 |0 |4120342 |
975
+ |aeb |Arabic |4120342 |0 |4120342 |
976
+ |afr |Afrikaans|4120342 |6108 |4126450 |
977
+ |ajp |Arabic |4120342 |0 |4120342 |
978
+ |als |Albanian|4120342 |0 |4120342 |
979
+ |amh |Amharic |4120342 |25327 |4145669 |
980
+ |apc |Arabic |4120342 |0 |4120342 |
981
+ |arb |Arabic |6424999 |216430 |6641429 |
982
+ |ars |Arabic |4120342 |0 |4120342 |
983
+ |ary |Arabic |4120342 |18076 |4138418 |
984
+ |arz |Arabic |4120342 |0 |4120342 |
985
+ |azb |Azerbaijani|4120342 |0 |4120342 |
986
+ |azj |Azerbaijani|4120342 |0 |4120342 |
987
+ |bel |Belarusian|4120342 |21273 |4141615 |
988
+ |ben |Bengali |4120342 |30661 |4151003 |
989
+ |bjn |Banjar |8240684 |2000 |8242684 |
990
+ |bul |Bulgarian|4120342 |37722 |4158064 |
991
+ |cat |Catalan |4120342 |66900 |4187242 |
992
+ |ceb |Cebuano |4120342 |0 |4120342 |
993
+ |ces |Czech |4120342 |179604 |4299946 |
994
+ |ckb |Kurdish |4120342 |0 |4120342 |
995
+ |cym |Welsh |4120342 |0 |4120342 |
996
+ |dan |Danish |4120342 |36310 |4156652 |
997
+ |deu |German |4120342 |1326722 |5447064 |
998
+ |ell |Greek |4120342 |40291 |4160633 |
999
+ |eng |English |9771427 |8066678 |17838105 |
1000
+ |epo |Esperanto|4120342 |0 |4120342 |
1001
+ |est |Estonian|4120342 |0 |4120342 |
1002
+ |eus |Basque |4120342 |0 |4120342 |
1003
+ |fin |Finnish |4120342 |457895 |4578237 |
1004
+ |fra |French |4120342 |835520 |4955862 |
1005
+ |gla |Scottish Gaelic|4120342 |0 |4120342 |
1006
+ |gle |Irish |4120342 |0 |4120342 |
1007
+ |glg |Galician|4120342 |0 |4120342 |
1008
+ |guj |Gujarati|4120342 |2157 |4122499 |
1009
+ |hat |Haitian Creole|4120342 |0 |4120342 |
1010
+ |hau |Hausa |4120342 |51396 |4171738 |
1011
+ |heb |Hebrew |4120342 |103466 |4223808 |
1012
+ |hin |Hindi |4120342 |260387 |4380729 |
1013
+ |hun |Hungarian|4120342 |82039 |4202381 |
1014
+ |hye |Armenian|4120342 |7080 |4127422 |
1015
+ |ibo |Igbo |4120342 |36312 |4156654 |
1016
+ |ind |Indonesian|4120342 |45709 |4166051 |
1017
+ |isl |Icelandic|4120342 |0 |4120342 |
1018
+ |ita |Italian |4120342 |405682 |4526024 |
1019
+ |jav |Javanese|4120342 |829 |4121171 |
1020
+ |jpn |Japanese|4120342 |2693177 |6813519 |
1021
+ |kan |Kannada |4120342 |1156 |4121498 |
1022
+ |kas |Kashmiri|4120342 |0 |4120342 |
1023
+ |kat |Georgian|4120342 |0 |4120342 |
1024
+ |kaz |Kazakh |4120342 |0 |4120342 |
1025
+ |khk |Mongolian|4120342 |0 |4120342 |
1026
+ |khm |Khmer |4120342 |0 |4120342 |
1027
+ |kir |Kyrgyz |4120342 |0 |4120342 |
1028
+ |kmr |Kurdish |4120342 |0 |4120342 |
1029
+ |knc |Kanuri |8240684 |0 |8240684 |
1030
+ |kor |Korean |4120342 |41011 |4161353 |
1031
+ |lao |Lao |4120342 |0 |4120342 |
1032
+ |lit |Lithuanian|4120342 |0 |4120342 |
1033
+ |ltz |Luxembourgish|4120342 |0 |4120342 |
1034
+ |lvs |Latvian |4120342 |0 |4120342 |
1035
+ |mal |Malayalam|4120342 |4347 |4124689 |
1036
+ |mar |Marathi |4120342 |3678 |4124020 |
1037
+ |min |Minangkabau|6753788 |2000 |6755788 |
1038
+ |mkd |Macedonian|4120342 |0 |4120342 |
1039
+ |mlt |Maltese |4120342 |0 |4120342 |
1040
+ |mni |Manipuri|4120342 |0 |4120342 |
1041
+ |mri |Maori |4120342 |0 |4120342 |
1042
+ |mya |Burmese |4120342 |0 |4120342 |
1043
+ |nld |Dutch |4120342 |220181 |4340523 |
1044
+ |nno |Norwegian|4120342 |0 |4120342 |
1045
+ |nob |Norwegian|4120342 |0 |4120342 |
1046
+ |npi |Nepali |4120342 |0 |4120342 |
1047
+ |nso |Northern Sotho|4120342 |0 |4120342 |
1048
+ |pbt |Pashto |4120342 |0 |4120342 |
1049
+ |pes |Persian |4120342 |245520 |4365862 |
1050
+ |plt |Malagasy|4120342 |0 |4120342 |
1051
+ |pol |Polish |4120342 |332503 |4452845 |
1052
+ |por |Portuguese|4120342 |287432 |4407774 |
1053
+ |ron |Romanian|4120342 |36359 |4156701 |
1054
+ |rus |Russian |4120342 |545920 |4666262 |
1055
+ |sin |Sinhala |4120342 |195 |4120537 |
1056
+ |slk |Slovak |4120342 |27845 |4148187 |
1057
+ |slv |Slovenian|4120342 |25731 |4146073 |
1058
+ |smo |Samoan |4120342 |0 |4120342 |
1059
+ |sna |Shona |4120342 |3684 |4124026 |
1060
+ |snd |Sindhi |4120342 |0 |4120342 |
1061
+ |som |Somali |4120342 |2926 |4123268 |
1062
+ |sot |Southern Sotho|4120342 |0 |4120342 |
1063
+ |spa |Spanish |4120342 |379194 |4499536 |
1064
+ |srp |Serbian |4120342 |77124 |4197466 |
1065
+ |sun |Sundanese|4120342 |2208 |4122550 |
1066
+ |swe |Swedish |4120342 |76486 |4196828 |
1067
+ |swh |Swahili |4120342 |12726 |4133068 |
1068
+ |tam |Tamil |4120342 |11462 |4131804 |
1069
+ |taq |Tamasheq|4120342 |0 |4120342 |
1070
+ |tel |Telugu |4120342 |477821 |4598163 |
1071
+ |tgk |Tajik |4120342 |0 |4120342 |
1072
+ |tha |Thai |4120342 |2125180 |6245522 |
1073
+ |tur |Turkish |4120342 |59932 |4180274 |
1074
+ |ukr |Ukrainian|4120342 |189384 |4309726 |
1075
+ |urd |Urdu |4120342 |337739 |4458081 |
1076
+ |uzn |Uzbek |4120342 |0 |4120342 |
1077
+ |vie |Vietnamese|4120342 |42232 |4162574 |
1078
+ |xho |Xhosa |4120342 |2952 |4123294 |
1079
+ |ydd |Yiddish |4120342 |0 |4120342 |
1080
+ |yor |Yoruba |4120342 |4907 |4125249 |
1081
+ |yue |Chinese |4120342 |0 |4120342 |
1082
+ |zho-Hans |Chinese |4120342 |54528 |4174870 |
1083
+ |zho-Hant |Chinese |4120342 |0 |4120342 |
1084
+ |zsm |Malay |4120342 |13950 |4134292 |
1085
+ |zul |Zulu |4120342 |786 |4121128 |
1086
+ |arq |Arabic |0 |6046 |6046 |
1087
+ |ban |Balinese|0 |2000 |2000 |
1088
+ |bbc |Toba Batak|0 |2000 |2000 |
1089
+ |bem |Bemba |0 |776 |776 |
1090
+ |fil |Filipino|0 |220 |220 |
1091
+ |fon |Fon |0 |845 |845 |
1092
+ |hrv |Croatian|0 |9007 |9007 |
1093
+ |kin |Kinyarwanda|0 |11165 |11165 |
1094
+ |lij |Ligurian|0 |6409 |6409 |
1095
+ |mad |Madurese|0 |2000 |2000 |
1096
+ |nij |Ngaju |0 |2000 |2000 |
1097
+ |nor |Norwegian|0 |72352 |72352 |
1098
+ |pan |Punjabi |0 |2156 |2156 |
1099
+ |twi |Twi |0 |10840 |10840 |
1100
+ |wol |Wolof |0 |785 |785 |
1101
+ |zho |Chinese |0 |74972 |74972 |
1102
+
1103
+ PS: Templated data also includes Mozambican Portuguese, which doesn't have its own ISO language code.
1104
+
1105
+ </details>
1106
+
1107
+ <br>
1108
+
1109
+
1110
+ # Motivations & Intentions
1111
+ - **Curation Rationale:** Automatic augmentation of existing datasets serves to enhance the available linguistic resources for multiple languages. The list of languages was initially established from mT5 and aligned with the annotators’ language list and NLLB translation model. The datasets were translated directly from English for all languages.
1112
+
1113
+
1114
+ # Additional Information
1115
+ ## Provenance
1116
+ - **Methods Used:** A combination of crowd-sourced templating and automatic translation was employed to source this dataset.
1117
+ - **Methodology Details:**
1118
+ - *Source:* Existing NLP datasets
1119
+ - *Dates of Collection:* May 2023 - Dec 2023
1120
+
1121
+
1122
+ ## Dataset Version and Maintenance
1123
+ - **Maintenance Status:** Actively Maintained
1124
+ - **Version Details:**
1125
+ - *Current version:* 1.0
1126
+ - *Last Update:* 02/2024
1127
+ - *First Release:* 02/2024
1128
+
1129
+
1130
+ ## Authorship
1131
+ - **Publishing Organization:** [Cohere For AI](https://cohere.com/research)
1132
+ - **Industry Type:** Not-for-profit - Tech
1133
+ - **Contact Details:** https://cohere.com/research/aya
1134
+
1135
+
1136
+ ## Licensing Information
1137
+ This dataset can be used for any purpose, whether academic or commercial, under the terms of the [Apache 2.0](https://opensource.org/license/apache-2-0) License.
1138
+
1139
+
1140
+ ## Citation Information
1141
+ ```bibtex
1142
+ @misc{singh2024aya,
1143
+ title={Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuning},
1144
+ author={Shivalika Singh and Freddie Vargus and Daniel Dsouza and Börje F. Karlsson and Abinaya Mahendiran and Wei-Yin Ko and Herumb Shandilya and Jay Patel and Deividas Mataciunas and Laura OMahony and Mike Zhang and Ramith Hettiarachchi and Joseph Wilson and Marina Machado and Luisa Souza Moura and Dominik Krzemiński and Hakimeh Fadaei and Irem Ergün and Ifeoma Okoh and Aisha Alaagib and Oshan Mudannayake and Zaid Alyafeai and Vu Minh Chien and Sebastian Ruder and Surya Guthikonda and Emad A. Alghamdi and Sebastian Gehrmann and Niklas Muennighoff and Max Bartolo and Julia Kreutzer and Ahmet Üstün and Marzieh Fadaee and Sara Hooker},
1145
+ year={2024},
1146
+ eprint={2402.06619},
1147
+ archivePrefix={arXiv},
1148
+ primaryClass={cs.CL}
1149
+ }
1150
+
1151
+ ```