Update README.md

README.md

@@ -2,4 +2,4 @@
 license: mit
 ---
 
-This dataset is a cleaned subset of the WikiText corpus for large language model pretraining. It has undergone preprocessing steps including cleaning, deduplication, normalization, and chunking to produce training-ready text segments. Due to its relatively small size, the dataset is intended for simple validation of text pretraining experiments rather than large-scale training.
+This dataset is a cleaned subset of the WikiText corpus (sourced from [Salesforce/wikitext](https://huggingface.co/datasets/Salesforce/wikitext)) for large language model pretraining. It has undergone preprocessing steps including cleaning, deduplication, normalization, and chunking to produce training-ready text segments. Due to its relatively small size, the dataset is intended for simple validation of text pretraining experiments rather than large-scale training.