Update README.md

README.md

@@ -6,7 +6,7 @@ tags: []
 # 5CD-AI/visocial-T5-base
 ## Overview
 <!-- Provide a quick summary of what the model is/does. -->
-We continually pretrain `google/mt5-base`[1] on a merged 20GB dataset, the training dataset includes:
+We trimmed vocabulary size to 50,589 and continually pretrained `google/mt5-base`[1] on a merged 20GB dataset, the training dataset includes:
 - Internal data (100M comments and 15M posts on Facebook)
 - UIT data[2], which is used to pretrain `uitnlp/visobert`[2]
 - MC4 ecommerce