suolyer commited on
Commit
59e661c
1 Parent(s): 8438491

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +48 -0
README.md CHANGED
@@ -1,3 +1,51 @@
1
  ---
 
 
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ language:
3
+ - zh
4
  license: apache-2.0
5
+
6
+ tags:
7
+ - bert
8
+ - NLU
9
+ - Sentiment
10
+
11
+ inference: true
12
+
13
+ widget:
14
+ - text: "今天心情不好"
15
+
16
  ---
17
+ # Erlangshen-MegatronBert-1.3B-Semtiment, model (Chinese),one model of [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM).
18
+ We collect 8 sentiment datasets in the Chinese domain for finetune, with a total of 227347 samples. Our model is mainly based on [MegatronBert-1.3B](https://huggingface.co/IDEA-CCNL/Erlangshen-MegatronBert-1.3B)
19
+
20
+ ## Usage
21
+ ```python
22
+ from transformers import BertForSequenceClassification
23
+ from transformers import BertTokenizer
24
+ import torch
25
+
26
+ tokenizer=BertTokenizer.from_pretrained('IDEA-CCNL/Erlangshen-MegatronBert-1.3B-Sentiment')
27
+ model=BertForSequenceClassification.from_pretrained('IDEA-CCNL/Erlangshen-MegatronBert-1.3B-Sentiment')
28
+
29
+ text='今天心情不好'
30
+
31
+ output=model(torch.tensor([tokenizer.encode(text)]))
32
+ print(torch.nn.functional.softmax(output.logits,dim=-1))
33
+
34
+ ```
35
+ ## Scores on downstream chinese tasks
36
+ | Model | ASAP-SENT | ASAP-ASPECT | ChnSentiCorp |
37
+ | :--------: | :-----: | :----: | :-----: |
38
+ | Erlangshen-Roberta-110M-Sentiment | 97.77 | 97.31 | 96.61 |
39
+ | Erlangshen-Roberta-330M-Sentiment | 97.9 | 97.51 | 96.66 |
40
+ | Erlangshen-MegatronBert-1.3B-Sentiment | 98.1 | 97.8 | 97 |
41
+
42
+ ## Citation
43
+ If you find the resource is useful, please cite the following website in your paper.
44
+ ```
45
+ @misc{Fengshenbang-LM,
46
+ title={Fengshenbang-LM},
47
+ author={IDEA-CCNL},
48
+ year={2021},
49
+ howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
50
+ }
51
+ ```