Jaemink commited on
Commit
4dfcdd4
1 Parent(s): 14787bf

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +59 -0
README.md ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # SentiCSE
2
+ This is a roBERTa-base model trained on MR dataset and finetuned for sentiment analysis with the Sentiment tasks.
3
+ This model is suitable for English.
4
+
5
+ - Reference Paper: SentiCSE (Main of Coling 2024).
6
+ - Git Repo: https://github.com/nayohan/SentiCSE.
7
+
8
+ ```python
9
+ import torch
10
+ from scipy.spatial.distance import cosine
11
+ from transformers import AutoTokenizer, AutoModel
12
+
13
+
14
+ tokenizer = AutoTokenizer.from_pretrained("DILAB-HYU/SentiCSE")
15
+ model = AutoModel.from_pretrained("DILAB-HYU/SentiCSE")
16
+
17
+ # Tokenize input texts
18
+ texts = [
19
+ "The food is delicious.",
20
+ "The atmosphere of the restaurant is good.",
21
+ "The food at the restaurant is devoid of flavor.",
22
+ "The restaurant lacks a good ambiance."
23
+ ]
24
+ inputs = tokenizer(texts, padding=True, truncation=True, return_tensors="pt")
25
+
26
+ # Get the embeddings
27
+ with torch.no_grad():
28
+ embeddings = model(**inputs, output_hidden_states=True, return_dict=True).pooler_output
29
+
30
+ # Calculate cosine similarities
31
+ # Cosine similarities are in [-1, 1]. Higher means more similar
32
+ cosine_sim_0_1 = 1 - cosine(embeddings[0], embeddings[1])
33
+ cosine_sim_0_2 = 1 - cosine(embeddings[0], embeddings[2])
34
+ cosine_sim_0_3 = 1 - cosine(embeddings[0], embeddings[3])
35
+
36
+ print("Cosine similarity between \"%s\" and \"%s\" is: %.3f" % (texts[0], texts[1], cosine_sim_0_1))
37
+ print("Cosine similarity between \"%s\" and \"%s\" is: %.3f" % (texts[0], texts[2], cosine_sim_0_2))
38
+ print("Cosine similarity between \"%s\" and \"%s\" is: %.3f" % (texts[0], texts[3], cosine_sim_0_3))
39
+
40
+ ```
41
+ Output:
42
+
43
+ ```
44
+ Cosine similarity between "The food is delicious." and "The atmosphere of the restaurant is good." is: 0.942
45
+ Cosine similarity between "The food is delicious." and "The food at the restaurant is devoid of flavor." is: 0.703
46
+ Cosine similarity between "The food is delicious." and "The restaurant lacks a good ambiance." is: 0.656
47
+ ```
48
+
49
+ ## BibTeX entry and citation info
50
+ Please cite the reference paper if you use this model.
51
+
52
+ ```
53
+ @article{2024SentiCES,
54
+ title={SentiCSE: A Sentiment-aware Contrastive Sentence Embedding Framework with Sentiment-guided Textual Similarity},
55
+ author={Kim, Jaemin and Na, Yohan and Kim, Kangmin and Lee, Sangrak and Chae, Dong-Kyu},
56
+ journal={Proceedings of the 30th International Conference on Computational Linguistics (COLING)},
57
+ year={2024},
58
+ }
59
+ ```