Update README.md
Browse files
README.md
CHANGED
@@ -457,7 +457,7 @@ PolyNews is a multilingual dataset containing news titles in 77 languages and 19
|
|
457 |
|
458 |
### Uses
|
459 |
|
460 |
-
This dataset can be used for domain adaptation of language models.
|
461 |
|
462 |
### Languages
|
463 |
|
@@ -580,7 +580,8 @@ For all languages, there is only the `train` split.
|
|
580 |
|
581 |
Multiple multilingual, human-translated, datasets containing news texts have been released in recent years.
|
582 |
However, these datasets are stored in different formats and various websites, and many contain numerous near duplicates.
|
583 |
-
With PolyNews, we aim to provide an easily-accessible, unified and deduplicated dataset that combines these disparate data sources
|
|
|
584 |
|
585 |
### Source Data
|
586 |
|
|
|
457 |
|
458 |
### Uses
|
459 |
|
460 |
+
This dataset can be used for domain adaptation of language models, language modeling or text generation.
|
461 |
|
462 |
### Languages
|
463 |
|
|
|
580 |
|
581 |
Multiple multilingual, human-translated, datasets containing news texts have been released in recent years.
|
582 |
However, these datasets are stored in different formats and various websites, and many contain numerous near duplicates.
|
583 |
+
With PolyNews, we aim to provide an easily-accessible, unified and deduplicated dataset that combines these disparate data sources.
|
584 |
+
It can be used for domain adaptation of language models, language modeling or text generation in both high-resource and low-resource languages.
|
585 |
|
586 |
### Source Data
|
587 |
|