Update references

README.md

@@ -7,12 +7,11 @@ widget:
 
 # Twitter-roBERTa-base for Sentiment Analysis - UPDATED (2021)
 
-This is a roBERTa-base model trained on ~124M tweets from January 2018 to December 2021 (see [TimeLMs paper](https://arxiv.org/abs/2202.03829)), and finetuned for sentiment analysis with the TweetEval benchmark. 
-The original roBERTa-base model can be found [here](https://huggingface.co/cardiffnlp/twitter-roberta-base-2021-124m). 
-This model is suitable for English. 
+This is a roBERTa-base model trained on ~124M tweets from January 2018 to December 2021 , and finetuned for sentiment analysis with the TweetEval benchmark. 
+The original roBERTa-base model can be found [here](https://huggingface.co/cardiffnlp/twitter-roberta-base-2021-124m) and the original reference paper is [TweetEval](https://github.com/cardiffnlp/tweeteval).This model is suitable for English. 
 
-- Reference Paper: [_TweetEval_ (Findings of EMNLP 2020)](https://arxiv.org/pdf/2010.12421.pdf). 
-- Git Repo: [Tweeteval official repository](https://github.com/cardiffnlp/tweeteval).
+- Reference Paper: [TimeLMs paper](https://arxiv.org/abs/2202.03829). 
+- Git Repo: [TimeLMs official repository](https://github.com/cardiffnlp/timelms).
 
 <b>Labels</b>: 
 0 -> Negative;