--- dataset_info: features: - name: text dtype: string - name: score dtype: float64 - name: embedding sequence: float64 - name: dataset dtype: string splits: - name: train num_bytes: 1199742546 num_examples: 110000 download_size: 856443525 dataset_size: 1199742546 configs: - config_name: default data_files: - split: train path: data/train-* license: apache-2.0 task_categories: - text-classification language: - pt tags: - portuguese - language-modeling pretty_name: GigaVerbo Text-Filter size_categories: - 100K ## Table of Contents - [Table of Contents](#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) - [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) - [Dataset Curators](#dataset-curators) - [Licensing Information](#licensing-information) - [Citation Information](#citation-information) - [Aknowlegments](#aknowlegments) - [Contributions](#contributions) ## Dataset Description - **Homepage:** https://huggingface.co/datasets/TucanoBR/GigaVerbo-Text-Filter - **Repository:** https://huggingface.co/datasets/TucanoBR/GigaVerbo-Text-Filter - **Paper:** [Tucano: Advancing Neural Text Generation for Portuguese](https://arxiv.org/abs/2411.07854) - **Point of Contact:** [Nk-correa](mailto:kluge@uni-bonn.de) ### Dataset Summary GigaVerbo Text-Filter is a dataset with 110,000 randomly selected samples from 9 subsets of [GigaVerbo](https://huggingface.co/datasets/TucanoBR/GigaVerbo) (i.e., specifically those that were not synthetic). This dataset was used to train the text-quality filters described in "_[Tucano: Advancing Neural Text Generation for Portuguese](https://arxiv.org/abs/2411.07854)_". To create the text embeddings, we used [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE). All scores were generated by GPT-4o. ### Supported Tasks and Leaderboards This dataset can be utilized for tasks involving text classification/regression in Portuguese. ### Languages Portuguese ## Dataset Structure ### Data Instances The dataset consists of the following features: - **text:** a string of text in Portuguese. - **score:** the score attributed by GPT-4o to that corresponding string of text. - **embedding:** embedding vector generated by [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE) to that corresponding string of text. - **name:** the subset of GigaVerbo from which the corresponding text string originated. ### Data Fields ```python { "text": "A inteligência artificial (de sigla: IA; do inglês: artificial intelligence, de sigla: AI) é um campo de estudo multidisciplinar que abrange varias áreas do conhecimento ...", "score": 0.85, "embedding": [0.313, 0.716, 0.897, 0.571, 0.061, 0.712, 0.265, 0.092, 0.816, 0.998, ...], "name" : "brwac" } ``` ### Data Splits Available splits are `train`. ```python from datasets import load_dataset dataset = load_dataset("TucanoBR/GigaVerbo-Text-Filter", split='train') # If you don't want to download the entire dataset, set streaming to `True` dataset = load_dataset("TucanoBR/GigaVerbo-Text-Filter", split='train', streaming=True) ``` ## Dataset Creation ### Curation Rationale This dataset was developed as part of the study "[Tucano: Advancing Neural Text Generation for Portuguese](https://arxiv.org/abs/2411.07854)". In short, GigaVerbo Text-Filter is a dataset with 110,000 randomly selected samples from 9 subsets of [GigaVerbo](https://huggingface.co/datasets/TucanoBR/GigaVerbo). ### Source Data #### Initial Data Collection and Normalization GigaVerbo Text-Filter has been scored GPT-4o. Text embeddings were generated by [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE). #### Who are the source language producers? All text samples are native to Portuguese or translated from other languages to Portuguese (slight contamination of different languages should also be expected). ### Annotations #### Annotation process GigaVerbo Text-Filter is a dataset with 110,000 randomly selected samples from 9 subsets of [GigaVerbo](https://huggingface.co/datasets/TucanoBR/GigaVerbo). All text samples are native to Portuguese or translated from other languages to Portuguese (slight contamination of different languages should also be expected). #### Who are the annotators? [Nicholas Kluge Corrêa](mailto:kluge@uni-bonn.de). ### Personal and Sensitive Information This dataset can potentially contain personal and sensitive information, along with offensive, toxic, and disturbing language. ## Considerations for Using the Data ### Social Impact of Dataset The presence of personal and sensitive information within the dataset raises concerns about privacy and data protection, potentially leading to breaches of individuals' confidentiality and security. Furthermore, the inclusion of offensive, toxic, and disturbing language in the dataset poses risks of perpetuating harmful behaviors and attitudes, contributing to the normalization of hate speech and online toxicity. Therefore, careful handling and ethical considerations are essential to mitigate these potential social impacts and promote responsible dataset use. ### Discussion of Biases The inclusion of offensive, toxic, and disturbing language in the dataset poses risks of perpetuating harmful behaviors and attitudes, contributing to the normalization of hate speech and online toxicity. ### Other Known Limitations A significant portion of the dataset's data has been translated using translation engines, potentially resulting in corrupted samples of both language and code. While useful for quickly converting text between languages, translation engines often struggle with accurately preserving the syntax, semantics, and context of programming languages. As a result, the translated code may contain errors, syntax inconsistencies, or even introduce vulnerabilities, rendering it unreliable or unusable for its intended purpose. ## Additional Information ### Dataset Curators [Nicholas Kluge Corrêa](mailto:kluge@uni-bonn.de). ### Licensing Information The following datasets and respective licenses from GigaVerbo (only training splits are a part of the corpus): - [HPLT-PT](https://huggingface.co/datasets/HPLT/hplt_monolingual_v1_2) (License: [cc0-1.0](https://huggingface.co/datasets/oscar-corpus/OSCAR-2301#licensing-information)) - [CC-2023](https://huggingface.co/datasets/dominguesm/CC-MAIN-2023-23) (License: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/deed.en)) - [CCc100](https://huggingface.co/datasets/eduagarcia/CrawlPT_dedup) (License: [Common Crawl terms of use](https://commoncrawl.org/terms-of-use/)) - [MC4-PT](https://huggingface.co/datasets/thegoodfellas/mc4-pt-cleaned) (License: [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)) - [Blogset-BR](https://huggingface.co/datasets/thegoodfellas/blogset-br) (License: [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)) - [BrWaC](https://huggingface.co/datasets/UFRGS/brwac) (License: Unknown) - [Wikipedia](https://huggingface.co/datasets/graelo/wikipedia) (License: [CC BY-SA 3.0](https://creativecommons.org/licenses/by-sa/3.0/)) - [Corpus Carolina](https://huggingface.co/datasets/carolina-c4ai/corpus-carolina) (License: [CC BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/deed.en)) - [CulturaX](https://huggingface.co/datasets/uonlp/CulturaX) (License: [ODC-By](https://opendatacommons.org/licenses/by/1-0/), [cc0-1.0](https://huggingface.co/datasets/oscar-corpus/OSCAR-2301#licensing-information)) - [OSCAR](https://huggingface.co/datasets/eduagarcia/CrawlPT_dedup) (License: [cc0-1.0](https://huggingface.co/datasets/oscar-corpus/OSCAR-2301#licensing-information)) - [Legal Portuguese](https://huggingface.co/datasets/eduagarcia/LegalPT_dedup) (License: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/deed.en)) - [Xlsum](https://huggingface.co/datasets/csebuetnlp/xlsum) (License: [CC BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/deed.en)) - [Roots Wikiquote](https://huggingface.co/datasets/bigscience-data/roots_pt_wikiquote) (License: [CC BY-SA 3.0](https://creativecommons.org/licenses/by-sa/3.0/)) - [Roots Ted Talks](https://huggingface.co/datasets/bigscience-data/roots_pt_ted_talks_iwslt) (License: [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en)) ### Citation Information ```latex @misc{correa2024tucanoadvancingneuraltext, title={{Tucano: Advancing Neural Text Generation for Portuguese}}, author={Corr{\^e}a, Nicholas Kluge and Sen, Aniket and Falk, Sophia and Fatimah, Shiza}, year={2024}, eprint={2411.07854}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2411.07854}, } ``` ### Aknowlegments We gratefully acknowledge the granted access to the [Marvin cluster](https://www.hpc.uni-bonn.de/en/systems/marvin) hosted by [University of Bonn](https://www.uni-bonn.de/en) along with the support provided by its High Performance Computing \& Analytics Lab. ### Contributions If you want to contribute, contact me at [kluge@uni-bonn.de](mailto:kluge@uni-bonn.de)!