Kostya165 commited on
Commit
8d10727
·
verified ·
1 Parent(s): ae44a1e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -53
README.md CHANGED
@@ -57,56 +57,3 @@ print(f"Предсказанный класс: {predicted_class}")
57
 
58
 
59
 
60
- ---
61
-
62
- ### English Version:
63
-
64
- ```markdown
65
- ---
66
- library_name: transformers
67
- tags: [emotion-detection, sentiment-analysis, lightweight]
68
- ---
69
-
70
- # Kostya165/rubert_emotion_slicer
71
-
72
- This is a fine-tuned version of the `blanchefort/rubert-base-cased-sentiment` model designed for emotion and sentiment analysis. The model recognizes five emotion classes: aggression, anxiety, sarcasm, positive, and neutral.
73
-
74
- ---
75
-
76
- ## Model Details
77
-
78
- - **Developed by:** Kostya165
79
- - **Model type:** BERT-based sequence classification
80
- - **Language:** Russian
81
- - **Finetuned from:** blanchefort/rubert-base-cased-sentiment
82
-
83
- ---
84
-
85
- ## Uses
86
-
87
- ### Direct Use
88
-
89
- The model can be used for sentiment and emotion analysis in Russian-language texts, such as chat messages, comments, or reviews.
90
-
91
- ### Out-of-Scope Use
92
-
93
- The model is not intended for analyzing texts in other languages or for determining complex emotional states beyond the five defined categories.
94
-
95
- ---
96
-
97
- ## How to Get Started
98
-
99
- To load and use the model, use the following code:
100
-
101
- ```python
102
- from transformers import BertTokenizer, BertForSequenceClassification
103
-
104
- model_name = "Kostya165/rubert_emotion_slicer"
105
- tokenizer = BertTokenizer.from_pretrained(model_name)
106
- model = BertForSequenceClassification.from_pretrained(model_name)
107
-
108
- text = "Мне очень понравилось!"
109
- inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True)
110
- outputs = model(**inputs)
111
- predicted_class = outputs.logits.argmax(dim=-1).item()
112
- print(f"Predicted class: {predicted_class}")
 
57
 
58
 
59