Update README.md

README.md

@@ -38,9 +38,6 @@ pretty_name: SPECTER
 - **Repository:** [More Information Needed](https://github.com/allenai/specter/blob/master/README.md)
 - **Paper:** [More Information Needed](https://arxiv.org/pdf/2004.07180.pdf)
 - **Point of Contact:** [@armancohan](https://github.com/armancohan), [@sergeyf](https://github.com/sergeyf), [@haroldrubio](https://github.com/haroldrubio), [@jinamshah](https://github.com/jinamshah)
-- **Size of downloaded dataset files:**
-- **Size of the generated dataset:** 
-- **Total amount of disk used:** 38.3 MB
 
 ### Dataset Summary
 
@@ -50,6 +47,9 @@ pretraining a Transformer language model on a powerful signal of document-level
 Unlike existing pretrained language models, SPECTER can be easily applied to 
 downstream applications without task-specific fine-tuning.
 
+Disclaimer: The team releasing SPECTER did not upload the dataset to the Hub and did not write a dataset card. 
+These steps were done by the Hugging Face team.
+
 ### Supported Tasks and Leaderboards
 
 [More Information Needed](https://github.com/allenai/specter)