Update README.md
Browse files
README.md
CHANGED
@@ -15,7 +15,6 @@ tags:
|
|
15 |
- emotion
|
16 |
- russian
|
17 |
- rubert
|
18 |
-
- tiny
|
19 |
- sentiment
|
20 |
- sentiment-analysis
|
21 |
- classification
|
@@ -25,7 +24,7 @@ datasets:
|
|
25 |
- cedr
|
26 |
---
|
27 |
|
28 |
-
This is [RuBERT
|
29 |
The task is a __multi-label classification__ with the following labels:
|
30 |
|
31 |
```yaml
|
@@ -52,9 +51,9 @@ anger: злость
|
|
52 |
|
53 |
```python
|
54 |
from transformers import pipeline
|
55 |
-
model = pipeline(model="seara/rubert-
|
56 |
model("Привет, ты мне нравишься!")
|
57 |
-
# [{'label': 'joy', 'score': 0.
|
58 |
```
|
59 |
|
60 |
## Dataset
|
@@ -68,20 +67,19 @@ An overview of the training data can be found in the source [article](https://ww
|
|
68 |
Training were done in this [project](https://github.com/searayeah/vkr-bert) with this parameters:
|
69 |
|
70 |
```yaml
|
71 |
-
tokenizer.max_length: null
|
72 |
batch_size: 64
|
73 |
optimizer: adam
|
74 |
lr: 0.00001
|
75 |
weight_decay: 0
|
76 |
-
num_epochs:
|
77 |
```
|
78 |
|
79 |
## Eval results (on test split)
|
80 |
|
81 |
| |no_emotion|joy |sadness|surprise|fear |anger|micro avg|macro avg|weighted avg|
|
82 |
|---------|----------|------|-------|--------|-------|-----|---------|---------|------------|
|
83 |
-
|precision|0.
|
84 |
-
|recall |0.84 |0.
|
85 |
-
|f1-score |0.
|
86 |
-
|auc-roc |0.
|
87 |
|support |734 |353 |379 |170 |141 |125 |1902 |1902 |1902 |
|
|
|
15 |
- emotion
|
16 |
- russian
|
17 |
- rubert
|
|
|
18 |
- sentiment
|
19 |
- sentiment-analysis
|
20 |
- classification
|
|
|
24 |
- cedr
|
25 |
---
|
26 |
|
27 |
+
This is [RuBERT](https://huggingface.co/DeepPavlov/rubert-base-cased) model fine-tuned for __emotion classification__ of short __Russian__ texts.
|
28 |
The task is a __multi-label classification__ with the following labels:
|
29 |
|
30 |
```yaml
|
|
|
51 |
|
52 |
```python
|
53 |
from transformers import pipeline
|
54 |
+
model = pipeline(model="seara/rubert-base-cased-cedr-russian-emotion")
|
55 |
model("Привет, ты мне нравишься!")
|
56 |
+
# [{'label': 'joy', 'score': 0.9388909935951233}]
|
57 |
```
|
58 |
|
59 |
## Dataset
|
|
|
67 |
Training were done in this [project](https://github.com/searayeah/vkr-bert) with this parameters:
|
68 |
|
69 |
```yaml
|
|
|
70 |
batch_size: 64
|
71 |
optimizer: adam
|
72 |
lr: 0.00001
|
73 |
weight_decay: 0
|
74 |
+
num_epochs: 5
|
75 |
```
|
76 |
|
77 |
## Eval results (on test split)
|
78 |
|
79 |
| |no_emotion|joy |sadness|surprise|fear |anger|micro avg|macro avg|weighted avg|
|
80 |
|---------|----------|------|-------|--------|-------|-----|---------|---------|------------|
|
81 |
+
|precision|0.87 |0.84 |0.85 |0.74 |0.7 |0.66 |0.83 |0.78 |0.83 |
|
82 |
+
|recall |0.84 |0.86 |0.82 |0.71 |0.74 |0.33 |0.79 |0.72 |0.79 |
|
83 |
+
|f1-score |0.86 |0.85 |0.84 |0.72 |0.72 |0.44 |0.81 |0.74 |0.8 |
|
84 |
+
|auc-roc |0.95 |0.97 |0.96 |0.94 |0.93 |0.86 |0.95 |0.93 |0.95 |
|
85 |
|support |734 |353 |379 |170 |141 |125 |1902 |1902 |1902 |
|