Datasets:
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: