Datasets:

habanoz
/

news-tr-1.8M

Modalities:

Formats:

Languages:

Size:

Libraries:

Dataset card Data Studio Files Files and versions Community

habanoz commited on Oct 6, 2024

Commit

3b0fa9c

·

verified ·

1 Parent(s): 3bef09d

Update README.md

Files changed (1) hide show

README.md +3 -2

README.md CHANGED Viewed

@@ -26,6 +26,7 @@ configs:
 News in the Turkish language. There are a total of 1.8M news articles and 2.1B words. Conversion to the number of tokens depends on the tokenizer used.
-- Collected from Common Crawl.
-- Filtered and deduplicated using Datatrove.

 News in the Turkish language. There are a total of 1.8M news articles and 2.1B words. Conversion to the number of tokens depends on the tokenizer used.
+I used cc_spider to collect the dataset.
+https://github.com/habanoz/cc_spider