Datasets:
metadata
language:
- pt
pretty_name: Aya Dataset Portuguese
tags:
- aya
- portuguese
- legal
- chemistry
license: apache-2.0
size_categories:
- 1K<n<10K
CohereForAI Aya Dataset filtrado para português (PT).
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