chiyuzhang commited on
Commit
fa2021d
1 Parent(s): c418377

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -0
README.md ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ tags:
7
+ - social media
8
+ - contrastive learning
9
+ ---
10
+ # Contrastive Learning of Sociopragmatic Meaning in Social Media
11
+
12
+ <p align="center"> <a href="https://chiyuzhang94.github.io/" target="_blank">Chiyu Zhang</a>, <a href="https://mageed.arts.ubc.ca/" target="_blank">Muhammad Abdul-Mageed</a>, <a href="https://ganeshjawahar.github.io/" target="_blank">Ganesh Jarwaha</a></p>
13
+ <p align="center" float="left">
14
+
15
+ <p align="center">Publish at Findings of ACL 2023</p>
16
+
17
+ [![Code License](https://img.shields.io/badge/Code%20License-Apache_2.0-green.svg)]()
18
+ [![Data License](https://img.shields.io/badge/Data%20License-CC%20By%20NC%204.0-red.svg)]()
19
+
20
+
21
+ <p align="center" width="100%">
22
+ <a><img src="https://github.com/UBC-NLP/infodcl/blob/master/images/infodcl_vis.png?raw=true" alt="Title" style="width: 90%; min-width: 300px; display: block; margin: auto;"></a>
23
+ </p>
24
+ Illustration of our proposed InfoDCL framework. We exploit distant/surrogate labels (i.e., emojis) to supervise two contrastive losses, corpus-aware contrastive loss (CCL) and Light label-aware contrastive loss (LCL-LiT). Sequence representations from our model should keep the cluster of each class distinguishable and preserve semantic relationships between classes.
25
+
26
+ ## Checkpoints of Models Pre-Trained with InfoDCL
27
+ * InfoDCL-RoBERTa trained with TweetEmoji-EN: https://huggingface.co/UBC-NLP/InfoDCL-emoji
28
+ * InfoDCL-RoBERTa trained with TweetHashtag-EN: https://huggingface.co/UBC-NLP/InfoDCL-hashtag
29
+
30
+ ## Model Performance
31
+
32
+ <p align="center" width="100%">
33
+ <a><img src="https://github.com/UBC-NLP/infodcl/blob/master/images/main_table.png?raw=true" alt="main table" style="width: 95%; min-width: 300px; display: block; margin: auto;"></a>
34
+ </p>
35
+ Fine-tuning results on our 24 Socio-pragmatic Meaning datasets (average macro-F1 over five runs).