Spaces:
Runtime error
Runtime error
Commit
•
dccce3b
1
Parent(s):
db866d2
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,12 +1,66 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
import os
|
|
|
|
| 3 |
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
|
| 8 |
def text_analysis(text):
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
html = '''<!doctype html>
|
| 11 |
<html>
|
| 12 |
<body>
|
|
@@ -29,8 +83,7 @@ def text_analysis(text):
|
|
| 29 |
</div>
|
| 30 |
</body>
|
| 31 |
</html>
|
| 32 |
-
'''.format(
|
| 33 |
-
|
| 34 |
return html
|
| 35 |
|
| 36 |
demo = gr.Interface(
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
import os
|
| 3 |
+
import torch
|
| 4 |
|
| 5 |
+
# Load model directly
|
| 6 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
# Install necessary libraries
|
| 10 |
+
!pip install transformers
|
| 11 |
+
!pip install torch
|
| 12 |
+
!pip install vaderSentiment
|
| 13 |
+
|
| 14 |
+
import torch
|
| 15 |
+
from transformers import RobertaTokenizer, RobertaForSequenceClassification
|
| 16 |
+
from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
|
| 17 |
+
|
| 18 |
+
# Load pre-trained RoBERTa model and tokenizer
|
| 19 |
+
tokenizer = RobertaTokenizer.from_pretrained('roberta-base')
|
| 20 |
+
model = RobertaForSequenceClassification.from_pretrained('roberta-base')
|
| 21 |
+
|
| 22 |
+
# Define a function to analyze text for potential adult content
|
| 23 |
+
def analyze_adult_content(text):
|
| 24 |
+
# Tokenize input text
|
| 25 |
+
inputs = tokenizer(text, return_tensors='pt')
|
| 26 |
+
|
| 27 |
+
# Perform inference
|
| 28 |
+
outputs = model(**inputs)
|
| 29 |
+
|
| 30 |
+
# Get predicted label (0: Not Adult Content, 1: Adult Content)
|
| 31 |
+
predicted_label_idx = torch.argmax(outputs.logits).item()
|
| 32 |
+
predicted_label = model.config.id2label[predicted_label_idx]
|
| 33 |
+
|
| 34 |
+
return predicted_label
|
| 35 |
+
|
| 36 |
+
# Define a function to analyze the sentiment of the text using VADER
|
| 37 |
+
def analyze_sentiment(text):
|
| 38 |
+
analyzer = SentimentIntensityAnalyzer()
|
| 39 |
+
sentiment_scores = analyzer.polarity_scores(text)
|
| 40 |
+
|
| 41 |
+
# Determine sentiment label based on compound score
|
| 42 |
+
if sentiment_scores['compound'] >= 0.05:
|
| 43 |
+
sentiment_label = 'Positive'
|
| 44 |
+
elif sentiment_scores['compound'] <= -0.05:
|
| 45 |
+
sentiment_label = 'Negative'
|
| 46 |
+
else:
|
| 47 |
+
sentiment_label = 'Neutral'
|
| 48 |
+
|
| 49 |
+
return sentiment_label, sentiment_scores
|
| 50 |
+
|
| 51 |
+
# Example text
|
| 52 |
+
text = "I really enjoy watching this movie, it's so entertaining!"
|
| 53 |
+
|
| 54 |
+
# Analyze adult content
|
| 55 |
+
adult_content_label = analyze_adult_content(text)
|
| 56 |
+
print("Adult Content Label:", adult_content_label)
|
| 57 |
|
| 58 |
def text_analysis(text):
|
| 59 |
+
# Analyze sentiment
|
| 60 |
+
sentiment_label, sentiment_scores = analyze_sentiment(text)
|
| 61 |
+
print("Sentiment Label:", sentiment_label)
|
| 62 |
+
print("Sentiment Scores:", sentiment_scores)
|
| 63 |
+
|
| 64 |
html = '''<!doctype html>
|
| 65 |
<html>
|
| 66 |
<body>
|
|
|
|
| 83 |
</div>
|
| 84 |
</body>
|
| 85 |
</html>
|
| 86 |
+
'''.format(sentiment_label, sentiment_scores, "Gamma", "Theta")
|
|
|
|
| 87 |
return html
|
| 88 |
|
| 89 |
demo = gr.Interface(
|