Upload 2 files
Browse files
app.py
ADDED
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# -*- coding: utf-8 -*-
|
2 |
+
"""app.ipynb
|
3 |
+
|
4 |
+
Automatically generated by Colab.
|
5 |
+
|
6 |
+
Original file is located at
|
7 |
+
https://colab.research.google.com/drive/1CbDOX8PDJB6ZyLZiLMXbPyr6k7dvrs20
|
8 |
+
"""
|
9 |
+
|
10 |
+
!pip install gradio
|
11 |
+
|
12 |
+
import gradio as gr
|
13 |
+
import torch
|
14 |
+
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
15 |
+
|
16 |
+
# Load the model and tokenizer
|
17 |
+
model_name = "qarib/bert-base-qarib"
|
18 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
19 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name, num_labels=2)
|
20 |
+
|
21 |
+
# Preprocessing function
|
22 |
+
def light_preprocess(text):
|
23 |
+
text = text.replace("@USER", "").replace("RT", "").strip()
|
24 |
+
return text
|
25 |
+
|
26 |
+
# Prediction function
|
27 |
+
def predict_offensive(text):
|
28 |
+
preprocessed_text = light_preprocess(text)
|
29 |
+
inputs = tokenizer(preprocessed_text, return_tensors="pt", truncation=True, padding=True)
|
30 |
+
with torch.no_grad():
|
31 |
+
outputs = model(**inputs)
|
32 |
+
logits = outputs.logits
|
33 |
+
predicted_class = torch.argmax(logits, dim=1).item()
|
34 |
+
return "Offensive" if predicted_class == 1 else "Not Offensive"
|
35 |
+
|
36 |
+
# Create the Gradio interface
|
37 |
+
iface = gr.Interface(
|
38 |
+
fn=predict_offensive,
|
39 |
+
inputs=gr.Textbox(lines=2, placeholder="Enter text here..."),
|
40 |
+
outputs="text",
|
41 |
+
title="Offensive Language Detection",
|
42 |
+
description="Enter a text to check if it's offensive or not.",
|
43 |
+
)
|
44 |
+
|
45 |
+
# Launch the interface
|
46 |
+
iface.launch()
|