Hugging Face's logo

t-bank-ai
/
response-toxicity-classifier-base

Text Classification
Transformers
PyTorch
Russian
bert
russian
pretraining
conversational
Inference Endpoints
Model card Files Community
1
amarkv commited on
Commit
7ade9e9
•
1 Parent(s): 7a2bac9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -5
README.md CHANGED
@@ -4,6 +4,13 @@ tags:
4
  - russian
5
  - pretraining
6
  license: mit
 
 
 
 
 
 
 
7
  ---
8
 
9
  # dialog-inapropriate-messages-classifier
@@ -25,9 +32,9 @@ The model was finetuned on DATASET_LINK.
25
 
26
  Model achieves the following results:
27
 
28
- | | OK - Precision | OK - Recall | OK - F1-score | TOXIC - Precision | TOXIC - Recall | TOXIC - F1-score | SEVERE TOXIC - Precision | SEVERE TOXIC - Recall | SEVERE TOXIC - F1-score | RISKS - Precision | RISKS - Recall | RISKS - F1-score |
29
- |-------------------------|----------------|-------------|---------------|-------------------|----------------|------------------|--------------------------|-----------------------|-------------------------|-------------------|----------------|------------------|
30
- | DATASET_TWITTER val.csv | 0.883 | 0.913 | 0.896 | 0.368 | 0.330 | 0.348 | 0.515 | 0.468 | 0.490 | 0.659 | 0.535 | 0.591 |
31
- | DATASET_GENA val.csv | 0.953 | 0.927 | 0.940 | 0.260 | 0.343 | 0.295 | 0.666 | 0.806 | 0.729 | 0.523 | 0.423 | 0.46 |
32
 
33
- The work was done during internship at Tinkoff by [Nikita Stepanov](https://huggingface.co/nikitast), mentored by [Alexander Markov](https://huggingface.co/amarkv).
 
4
  - russian
5
  - pretraining
6
  license: mit
7
+ widget:
8
+ - text: "[CLS] привет [SEP] привет! [SEP] как дела? [RESPONSE_TOKEN] норм"
9
+ example_title: "Dialog example 1"
10
+ - text: "[CLS] привет [SEP] привет! [SEP] как дела? [RESPONSE_TOKEN] соси вола"
11
+ example_title: "Dialog example 2"
12
+ - text: "[CLS] здравствуйте товарищ [RESPONSE_TOKEN] что это за говно на тебе надето?))"
13
+ example_title: "Dialog example 3"
14
  ---
15
 
16
  # dialog-inapropriate-messages-classifier
 
32
 
33
  Model achieves the following results:
34
 
35
+ | | OK - F1-score | TOXIC - F1-score | SEVERE TOXIC - F1-score | RISKS - F1-score |
36
+ |-------------------------|-------------------------|-------------------|----------------|------------------|
37
+ | DATASET_TWITTER val.csv | 0.896 | 0.348 | 0.490 | 0.591 |
38
+ | DATASET_GENA val.csv | 0.940 | 0.295 | 0.729 | 0.46 |
39
 
40
+ The work was done during internship at Tinkoff by [Nikita Stepanov](https://huggingface.co/nikitast).