sagorsarker
/

news_article_doc2vec

Model card Files Files and versions Community

sagorsarker commited on Nov 9, 2022

Commit

efb9cf2

•

1 Parent(s): 07ac9cf

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -6,6 +6,6 @@ News article doc2vec model trained for [bnlp](https://github.com/sagorbrur/bnlp)
 ## Training details
 - Trained using gensim doc2vec module
-- Trained on https://www.kaggle.com/datasets/ebiswas/bangla-largest-newspaper-dataset datasets with only three json for RAM limitations
-- Total news articles 91716
 - Parameter: vector size=100, min_count = 2 , epochs=40

 ## Training details
 - Trained using gensim doc2vec module
+- Trained on https://www.kaggle.com/datasets/ebiswas/bangla-largest-newspaper-dataset datasets with 8 jsons (randomly selected)
+- Total news articles 400013
 - Parameter: vector size=100, min_count = 2 , epochs=40