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: