Datasets:
File size: 701 Bytes
f77496e 8367e31 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
[Aya Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) filtrado para português.
**Aya Dataset Summary**
The [Aya Dataset](https://huggingface.co/datasets/CohereForAI/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: |