Update README.md

README.md

@@ -8,8 +8,12 @@ license: mit
 
 # Dataset Card for HLSNKIOPUS100
 
-This dataset contains English-Persian translation pairs from the OPUS-100 corpus. This corpus is noisy and has problematic entries.
-We're working on a clean corpus of this dataset.
+This dataset contains English-Persian translation pairs from the OPUS-100 corpus. 
+The original corpus of OPUS-100 is movie dialogues and gets repetitive really fast, 
+it also contains duplicates, false translations and huge chunks of unrelated data, 
+making the dataset very noisy and almost unsuitable for any usage. 
+We have deduplicated, denoised and cleaned the dataset, reviewed the translations, 
+retranslated and cleared the meaning across all the entries to reach a standard version suitable for deep learning tasks.
 
 ## Dataset Details