aya_dataset_pt / README.md
nicolasdec's picture
Update README.md
f77496e verified
|
raw
history blame
701 Bytes

Aya Dataset filtrado para português.

Aya Dataset Summary

The Aya Dataset is a multilingual instruction fine-tuning dataset curated by an open-science community via Aya Annotation Platform from Cohere For AI. The dataset contains a total of 204k human-annotated prompt-completion pairs along with the demographics data of the annotators. This dataset can be used to train, finetune, and evaluate multilingual LLMs.

Curated by: Contributors of Aya Open Science Intiative.

Language(s): 65 languages (71 including dialects & scripts).

License: Apache 2.0

Aya Datasets Family: