Update README.md

README.md

@@ -12,9 +12,8 @@ model-index:
 - name: bertweet-base-finetuned-SARC-DS
   results: []
 ---
-
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
+# Utilising Weak Supervision to Create S3D: A Sarcasm Annotated Dataset
+This is the repository for the S3D dataset published at EMNLP 2022. The dataset can help build sarcasm detection models.
 
 # bertweet-base-finetuned-SARC-DS
 
@@ -28,16 +27,13 @@ It achieves the following results on the evaluation set:
 
 ## Model description
 
-The given description for BERTweet by VinAI is as follows:
+The given description for BERTweet by VinAI is as follows: <br>
 BERTweet is the first public large-scale language model pre-trained for English Tweets. BERTweet is trained based on the RoBERTa pre-training procedure. The corpus used to pre-train BERTweet consists of 850M English Tweets (16B word tokens ~ 80GB), containing 845M Tweets streamed from 01/2012 to 08/2019 and 5M Tweets related to the COVID-19 pandemic.
-
-## Intended uses & limitations
-
-More information needed
+<br>
 
 ## Training and evaluation data
 
-More information needed
+This [vinai/bertweet-base](https://huggingface.co/vinai/bertweet-base) model was finetuned on the [SARC](https://metatext.io/datasets/self-annotated-reddit-corpus-(sarc)) dataset. The dataset is intended to help build sarcasm detection models.
 
 ## Training procedure