nicolasdec commited on
Commit
8367e31
1 Parent(s): 941a44c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +43 -0
README.md CHANGED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ license: apache-2.0
2
+ dataset_info:
3
+ - config_name: default
4
+ features:
5
+ - name: inputs
6
+ dtype: string
7
+ - name: targets
8
+ dtype: string
9
+ - name: language
10
+ dtype: string
11
+ - name: language_code
12
+ dtype: string
13
+ _type: Value
14
+ - name: annotation_type
15
+ dtype: string
16
+ - name: user_id
17
+ dtype: string
18
+ annotations_creators:
19
+ - crowdsourced
20
+ - expert-generated
21
+ language:
22
+ - por
23
+ language_creators:
24
+ - crowdsourced
25
+ - expert-generated
26
+ multilinguality:
27
+ - multilingual
28
+ pretty_name: Aya Dataset Portuguese
29
+
30
+ [Aya Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) filtrado para poruguese.
31
+
32
+ **Aya Dataset Summary**
33
+
34
+ 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.
35
+ This dataset can be used to train, finetune, and evaluate multilingual LLMs.
36
+
37
+ Curated by: Contributors of Aya Open Science Intiative.
38
+
39
+ Language(s): 65 languages (71 including dialects & scripts).
40
+
41
+ License: Apache 2.0
42
+
43
+ Aya Datasets Family: