Spaces:
Running
Running
Update emotion_detection.py
Browse files- emotion_detection.py +10 -24
emotion_detection.py
CHANGED
@@ -3,6 +3,7 @@ from transformers_interpret import SequenceClassificationExplainer
|
|
3 |
import torch
|
4 |
import pandas as pd
|
5 |
|
|
|
6 |
class EmotionDetection:
|
7 |
"""
|
8 |
Emotion Detection on text data.
|
@@ -24,10 +25,12 @@ class EmotionDetection:
|
|
24 |
Parameters:
|
25 |
text (str): The user input string to emotion justification
|
26 |
Returns:
|
27 |
-
html (
|
28 |
"""
|
|
|
29 |
word_attributions = self.explainer(text)
|
30 |
html = self.explainer.visualize("example.html")
|
|
|
31 |
return html
|
32 |
|
33 |
def classify(self, text):
|
@@ -43,28 +46,9 @@ class EmotionDetection:
|
|
43 |
outputs = self.model(**tokens)
|
44 |
probs = torch.nn.functional.softmax(outputs[0], dim=-1)
|
45 |
probs = probs.mean(dim=0).detach().numpy()
|
|
|
|
|
46 |
|
47 |
-
# Original labels from model
|
48 |
-
original_labels = list(self.model.config.id2label.values())
|
49 |
-
|
50 |
-
# Only keep the 4 specific emotions and map custom names
|
51 |
-
desired_labels = ['joy', 'anger', 'sadness', 'optimism']
|
52 |
-
custom_labels = {
|
53 |
-
'joy': 'Happiness π',
|
54 |
-
'anger': 'Anger π‘',
|
55 |
-
'sadness': 'Sadness π’',
|
56 |
-
'optimism': 'Hopeful β¨'
|
57 |
-
}
|
58 |
-
|
59 |
-
filtered_probs = []
|
60 |
-
filtered_labels = []
|
61 |
-
|
62 |
-
for label, prob in zip(original_labels, probs):
|
63 |
-
if label in desired_labels:
|
64 |
-
filtered_probs.append(prob)
|
65 |
-
filtered_labels.append(custom_labels[label])
|
66 |
-
|
67 |
-
preds = pd.Series(filtered_probs, index=filtered_labels, name='Predicted Probability')
|
68 |
return preds
|
69 |
|
70 |
def run(self, text):
|
@@ -74,8 +58,10 @@ class EmotionDetection:
|
|
74 |
text (str): The user input string to perform emotion classification on
|
75 |
Returns:
|
76 |
predictions (str): The predicted probabilities for emotion classes
|
77 |
-
html (
|
78 |
"""
|
|
|
79 |
preds = self.classify(text)
|
80 |
html = self.justify(text)
|
81 |
-
|
|
|
|
3 |
import torch
|
4 |
import pandas as pd
|
5 |
|
6 |
+
|
7 |
class EmotionDetection:
|
8 |
"""
|
9 |
Emotion Detection on text data.
|
|
|
25 |
Parameters:
|
26 |
text (str): The user input string to emotion justification
|
27 |
Returns:
|
28 |
+
html (hmtl): html object for plotting emotion prediction justification
|
29 |
"""
|
30 |
+
|
31 |
word_attributions = self.explainer(text)
|
32 |
html = self.explainer.visualize("example.html")
|
33 |
+
|
34 |
return html
|
35 |
|
36 |
def classify(self, text):
|
|
|
46 |
outputs = self.model(**tokens)
|
47 |
probs = torch.nn.functional.softmax(outputs[0], dim=-1)
|
48 |
probs = probs.mean(dim=0).detach().numpy()
|
49 |
+
labels = list(self.model.config.id2label.values())
|
50 |
+
preds = pd.Series(probs, index=labels, name='Predicted Probability')
|
51 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
return preds
|
53 |
|
54 |
def run(self, text):
|
|
|
58 |
text (str): The user input string to perform emotion classification on
|
59 |
Returns:
|
60 |
predictions (str): The predicted probabilities for emotion classes
|
61 |
+
html (hmtl): html object for plotting emotion prediction justification
|
62 |
"""
|
63 |
+
|
64 |
preds = self.classify(text)
|
65 |
html = self.justify(text)
|
66 |
+
|
67 |
+
return preds, html
|