Update README.md
Browse files
README.md
CHANGED
@@ -11,26 +11,6 @@ widget:
|
|
11 |
```python
|
12 |
|
13 |
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
14 |
-
LABEL_COLUMNS = ['Self-direction: thought',
|
15 |
-
'Self-direction: action',
|
16 |
-
'Stimulation',
|
17 |
-
'Hedonism',
|
18 |
-
'Achievement',
|
19 |
-
'Power: dominance',
|
20 |
-
'Power: resources',
|
21 |
-
'Face',
|
22 |
-
'Security: personal',
|
23 |
-
'Security: societal',
|
24 |
-
'Tradition',
|
25 |
-
'Conformity: rules',
|
26 |
-
'Conformity: interpersonal',
|
27 |
-
'Humility',
|
28 |
-
'Benevolence: caring',
|
29 |
-
'Benevolence: dependability',
|
30 |
-
'Universalism: concern',
|
31 |
-
'Universalism: nature',
|
32 |
-
'Universalism: tolerance',
|
33 |
-
'Universalism: objectivity']
|
34 |
|
35 |
tokenizer = AutoTokenizer.from_pretrained("tum-nlp/Deberta_Human_Value_Detector")
|
36 |
model = AutoModelForSequenceClassification.from_pretrained("tum-nlp/Deberta_Human_Value_Detector", trust_remote_code=True)
|
@@ -51,6 +31,15 @@ with torch.no_grad():
|
|
51 |
test_prediction = trained_model(encoding["input_ids"], encoding["attention_mask"])
|
52 |
test_prediction = test_prediction["logits"].flatten().numpy()
|
53 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
54 |
print(f"Predictions:")
|
55 |
for label, prediction in zip(LABEL_COLUMNS, test_prediction):
|
56 |
if prediction < THRESHOLD:
|
|
|
11 |
```python
|
12 |
|
13 |
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
tokenizer = AutoTokenizer.from_pretrained("tum-nlp/Deberta_Human_Value_Detector")
|
16 |
model = AutoModelForSequenceClassification.from_pretrained("tum-nlp/Deberta_Human_Value_Detector", trust_remote_code=True)
|
|
|
31 |
test_prediction = trained_model(encoding["input_ids"], encoding["attention_mask"])
|
32 |
test_prediction = test_prediction["logits"].flatten().numpy()
|
33 |
|
34 |
+
```
|
35 |
+
|
36 |
+
## Prediction
|
37 |
+
To make a prediction and map the the outputs to the correct labels.
|
38 |
+
During the competiton a threshold of 0.25 was used to binarize the output.
|
39 |
+
```
|
40 |
+
THRESHOLD = 0.25
|
41 |
+
LABEL_COLUMNS = ['Self-direction: thought','Self-direction: action','Stimulation','Hedonism','Achievement','Power: dominance','Power: resources','Face','Security: personal',
|
42 |
+
'Security: societal','Tradition','Conformity: rules','Conformity: interpersonal','Humility','Benevolence: caring','Benevolence: dependability','Universalism: concern','Universalism: nature','Universalism: tolerance','Universalism: objectivity']
|
43 |
print(f"Predictions:")
|
44 |
for label, prediction in zip(LABEL_COLUMNS, test_prediction):
|
45 |
if prediction < THRESHOLD:
|