Update README.md

README.md

@@ -10,27 +10,16 @@ pretty_name: ur_para
 size_categories:
 - 100K<n<1M
 ---
-# Dataset Name
-ur-para (Urdu Paraphrasing Dataset) 
+# Paraphrase Dataset (Urdu)
 
-## Dataset Size
-329 MB
+This dataset contains paraphrases in Urdu. It is provided in the Parquet format and is split into a training set with 393,000 rows.
 
-## Dataset Description
-This dataset consists of pairs of Urdu sentences that have the same meaning but different phrasing. It is intended for use in natural language processing tasks such as text classification, machine translation, and semantic similarity analysis.
+## Dataset Details
 
-## Dataset Source
-Private collection
+- Columns:
+  - `sentence1`: The first sentence in a pair of paraphrases (string).
+  - `sentence2`: The second sentence in a pair of paraphrases (string).
 
-## Number of Samples
-n < 1k (exact number not disclosed due to privacy concerns)
+## Usage
 
-## Features:
-sentence1 (string): the original Urdu sentence
-sentence2 (string): the paraphrased Urdu sentence with the same meaning as sentence1
-
-## Format
-JSON
-
-### Note
-The sentences in this dataset are taken from various sources and may contain regional variations in the Urdu language.
+You can use this dataset for various natural language processing tasks such as text similarity, paraphrase identification, and language generation.