egoriya commited on
Commit
b772afd
1 Parent(s): a02a899

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +39 -0
README.md CHANGED
@@ -1,3 +1,42 @@
1
  ---
2
  license: mit
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
  ---
4
+
5
+ This classification model is based on [sberbank-ai/ruRoberta-large](https://huggingface.co/sberbank-ai/ruRoberta-large).
6
+ The model should be used to produce relevance and specificity of the last message in the context of a dialog.
7
+
8
+ It is pretrained on corpus of dialog data from social networks and finetuned on [tinkoff-ai/context_similarity](https://huggingface.co/tinkoff-ai/context_similarity).
9
+ The performance of the model on validation split [tinkoff-ai/context_similarity](https://huggingface.co/tinkoff-ai/context_similarity) (with the best thresholds for validation samples):
10
+
11
+ <table>
12
+ <thead>
13
+ <tr>
14
+ <td colspan="2">relevance</td>
15
+ <td colspan="2">specificity</td>
16
+ </tr>
17
+ </thead>
18
+ <tbody>
19
+ <tr>
20
+ <td>f0.5</td>
21
+ <td>roc-auc</td>
22
+ <td>f0.5</td>
23
+ <td>roc-auc</td>
24
+ </tr>
25
+ <tr>
26
+ <td>0.86</td>
27
+ <td>0.83</td>
28
+ <td>0.85</td>
29
+ <td>0.86</td>
30
+ </tr>
31
+ </tbody>
32
+ </table>
33
+
34
+ The model can be loaded as follows:
35
+
36
+ ```python
37
+ # pip install transformers
38
+ from transformers import AutoTokenizer, AutoModel
39
+ tokenizer = AutoTokenizer.from_pretrained("tinkoff-ai/context_similarity")
40
+ model = AutoModel.from_pretrained("tinkoff-ai/context_similarity")
41
+ # model.cuda()
42
+ ```