Update README.md

README.md

@@ -20,7 +20,7 @@ We offer datasets with **structured, high-quality, and continuously updated** da
 A large-scale transcribed speech dataset sourced from screen content, suitable for ASR and NLP tasks.
 
 - **XS Size** – Limited sample dataset.
-- ~~**S / M / L Size** – Get access to 20%, 50%, or 80% of the dataset with periodic updates.~~
+- **S Size** – Get access to 25% of the dataset with periodic updates.
 - **Full Size** – Full access + real-time updates.
 
 👉 [Explore ScreenTalk Dataset](https://huggingface.co/datasets/DataLabX/ScreenTalk-XS)