Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,43 @@
|
|
1 |
-
---
|
2 |
-
license: apache-2.0
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
---
|
4 |
+
|
5 |
+
```python
|
6 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
7 |
+
import numpy as np
|
8 |
+
|
9 |
+
def sigmoid(x):
|
10 |
+
return 1/(1 + np.exp(-x))
|
11 |
+
|
12 |
+
classes = [
|
13 |
+
'code_design',
|
14 |
+
'code_style',
|
15 |
+
'code_naming',
|
16 |
+
'code_logic',
|
17 |
+
'code_io',
|
18 |
+
'code_data',
|
19 |
+
'code_doc',
|
20 |
+
'code_api',
|
21 |
+
'compatibility',
|
22 |
+
'rule_def',
|
23 |
+
'config_commit_patch_review',
|
24 |
+
'config_building_installing',
|
25 |
+
]
|
26 |
+
class2id = {class_:id for id, class_ in enumerate(classes)}
|
27 |
+
id2class = {id:class_ for class_, id in class2id.items()}
|
28 |
+
|
29 |
+
checkpoint = 'mochodek/bert4comment-subject'
|
30 |
+
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
|
31 |
+
model = AutoModelForSequenceClassification.from_pretrained(checkpoint)
|
32 |
+
|
33 |
+
text = "What do you think about making this constant?"
|
34 |
+
encoded_input = tokenizer(text, return_tensors='pt')
|
35 |
+
output = model(**encoded_input)
|
36 |
+
|
37 |
+
logits = output.logits.detach().numpy()
|
38 |
+
|
39 |
+
scores = sigmoid(logits)
|
40 |
+
scores = (scores > 0.5).astype(int).reshape(-1)
|
41 |
+
scores_labels = [class_name for class_name in classes if scores[class2id[class_name]] == 1 ]
|
42 |
+
|
43 |
+
```
|