Update README.md

README.md

@@ -21,9 +21,4 @@ configs:
 ---
 
 ## Dataset Creation Process
-These subsets were created by streaming over the rows from the original dataset and filtering by the meta column. Each subset is generally limited to the first 100,000 qualifying rows encountered.
-
-**Note**: Some subsets contain fewer than 100,000 rows due to the sparsity of certain categories in the original dataset. In these cases, the rows may not be the first encountered but are unevenly distributed due to parallelization during collection.
-
-## Future Updates
-Future updates may include completing the smaller subsets to reach the 100,000 row target, which could result in breaking changes for those specific subsets.
+These subsets were created by streaming over the rows from [`monology/pile-uncopyrighted`](https://huggingface.co/datasets/monology/pile-uncopyrighted) and filtering by the meta column. Each subset is generally limited to the first 100,000 qualifying rows encountered.