Update README.md

README.md

@@ -18,4 +18,4 @@ configs:
 
 This dataset consists of a mixture of [TED talk transcripts](https://www.kaggle.com/datasets/miguelcorraljr/ted-ultimate-dataset), the high-quality [Aeon essays](https://www.kaggle.com/datasets/mannacharya/aeon-essays-dataset), as well as a filtered selection of [NPR Interview Transcripts](https://www.kaggle.com/datasets/shuyangli94/interview-npr-media-dialog-transcripts).
 
-All NPR transcripts were filtered to contain only the turns with > 50 words.
+All NPR transcripts were filtered to contain only the turns with > 100 words.