Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
from transformers import BertTokenizerFast, BertForTokenClassification
|
3 |
+
import gradio as gr
|
4 |
+
|
5 |
+
# Load tokenizer and model
|
6 |
+
tokenizer = BertTokenizerFast.from_pretrained('bert-base-uncased')
|
7 |
+
model = BertForTokenClassification.from_pretrained('./saved_model3')
|
8 |
+
model.eval()
|
9 |
+
model.to('cuda' if torch.cuda.is_available() else 'cpu')
|
10 |
+
|
11 |
+
# Define label mappings
|
12 |
+
id2label = {
|
13 |
+
0: 'O',
|
14 |
+
1: 'B-STEREO',
|
15 |
+
2: 'I-STEREO',
|
16 |
+
3: 'B-GEN',
|
17 |
+
4: 'I-GEN',
|
18 |
+
5: 'B-UNFAIR',
|
19 |
+
6: 'I-UNFAIR'
|
20 |
+
}
|
21 |
+
|
22 |
+
def predict_ner_tags(sentence):
|
23 |
+
inputs = tokenizer(sentence, return_tensors="pt", padding=True, truncation=True, max_length=128)
|
24 |
+
input_ids = inputs['input_ids'].to(model.device)
|
25 |
+
attention_mask = inputs['attention_mask'].to(model.device)
|
26 |
+
|
27 |
+
with torch.no_grad():
|
28 |
+
outputs = model(input_ids=input_ids, attention_mask=attention_mask)
|
29 |
+
logits = outputs.logits
|
30 |
+
probabilities = torch.sigmoid(logits)
|
31 |
+
predicted_labels = (probabilities > 0.5).int()
|
32 |
+
|
33 |
+
result = []
|
34 |
+
tokens = tokenizer.convert_ids_to_tokens(input_ids[0])
|
35 |
+
for i, token in enumerate(tokens):
|
36 |
+
if token not in tokenizer.all_special_tokens:
|
37 |
+
label_indices = (predicted_labels[0][i] == 1).nonzero(as_tuple=False).squeeze(-1)
|
38 |
+
labels = [id2label[idx.item()] for idx in label_indices] if label_indices.numel() > 0 else ['O']
|
39 |
+
result.append((token, labels))
|
40 |
+
|
41 |
+
return result
|
42 |
+
|
43 |
+
def format_output(result):
|
44 |
+
formatted_output = ""
|
45 |
+
for token, labels in result:
|
46 |
+
formatted_output += f"{token}: {', '.join(labels)}\n"
|
47 |
+
return formatted_output
|
48 |
+
|
49 |
+
iface = gr.Interface(
|
50 |
+
fn=predict_ner_tags,
|
51 |
+
inputs="text",
|
52 |
+
outputs="text",
|
53 |
+
title="Named Entity Recognition with BERT",
|
54 |
+
description="Enter a sentence to predict NER tags using BERT model trained for multi-label classification.",
|
55 |
+
examples=["Tall men are so clumsy."],
|
56 |
+
allow_flagging="never",
|
57 |
+
interpretation="default",
|
58 |
+
postprocessing_fn=format_output
|
59 |
+
)
|
60 |
+
|
61 |
+
if __name__ == "__main__":
|
62 |
+
iface.launch()
|