Update README.md

README.md

@@ -29,13 +29,11 @@ language:
 tags:
 - Japanese
 - good dataset
-pretty_name: 'Audio and text. Japanese. Edited for NLP and ASR training. Clean and accurate.'
+pretty_name: 'Audio and text. Japanese.'
 size_categories:
 - 100K<n<1M
 ---
 
--Each sample has been evaluated for accuracy.
-
 -Noise and silence have been removed from the begining and end of each sample. 
 
 -Unnecessary and inaccurate punctuation have been removed. 
@@ -43,7 +41,6 @@ size_categories:
 -Text has been normalized.
   Normalization is based on the neologd's rules: https://github.com/neologd/mecab-ipadic-neologd/wiki/Regexp.ja.
 
--Sample length in seconds for each audio sample has been added to the dataset for arbitrary token mapping.