Convert dataset sizes from base 2 to base 10 in the dataset card (#1)

- Convert dataset sizes from base 2 to base 10 in the dataset card (9f17d5e1a3c19d570fc7cb9e037f849e56ca43b3)

README.md

@@ -74,9 +74,9 @@ dataset_info:
 - **Repository:** https://github.com/google-research-datasets/wiki-split
 - **Paper:** [Learning To Split and Rephrase From Wikipedia Edit History](https://arxiv.org/abs/1808.09468)
 - **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
-- **Size of downloaded dataset files:** 95.63 MB
-- **Size of the generated dataset:** 370.41 MB
-- **Total amount of disk used:** 466.04 MB
+- **Size of downloaded dataset files:** 100.28 MB
+- **Size of the generated dataset:** 388.40 MB
+- **Total amount of disk used:** 488.68 MB
 
 ### Dataset Summary
 
@@ -98,9 +98,9 @@ the dataset contains some inherent noise, it can serve as valuable training data
 
 #### default
 
-- **Size of downloaded dataset files:** 95.63 MB
-- **Size of the generated dataset:** 370.41 MB
-- **Total amount of disk used:** 466.04 MB
+- **Size of downloaded dataset files:** 100.28 MB
+- **Size of the generated dataset:** 388.40 MB
+- **Total amount of disk used:** 488.68 MB
 
 An example of 'train' looks as follows.
 ```