Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
from transformers import AutoModelWithHeads, AutoTokenizer
|
3 |
+
import torch
|
4 |
+
|
5 |
+
# Load pre-trained BERT model with adapter support
|
6 |
+
st.title("Adapter Transformers for Text Classification")
|
7 |
+
|
8 |
+
@st.cache_resource
|
9 |
+
def load_model():
|
10 |
+
model = AutoModelWithHeads.from_pretrained("bert-base-uncased")
|
11 |
+
tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
|
12 |
+
|
13 |
+
# Add and activate an adapter
|
14 |
+
adapter_name = "my_adapter"
|
15 |
+
model.add_adapter(adapter_name)
|
16 |
+
model.train_adapter(adapter_name)
|
17 |
+
model.set_active_adapters(adapter_name)
|
18 |
+
|
19 |
+
# Add a classification head (binary classification)
|
20 |
+
model.add_classification_head(adapter_name, num_labels=2)
|
21 |
+
return model, tokenizer
|
22 |
+
|
23 |
+
# Load the model
|
24 |
+
model, tokenizer = load_model()
|
25 |
+
|
26 |
+
# Streamlit input
|
27 |
+
input_text = st.text_input("Enter text for classification:", "Steve Jobs founded Apple")
|
28 |
+
|
29 |
+
if input_text:
|
30 |
+
# Tokenize the input
|
31 |
+
inputs = tokenizer(input_text, return_tensors="pt", truncation=True, padding=True, max_length=512)
|
32 |
+
|
33 |
+
# Make the prediction
|
34 |
+
with torch.no_grad():
|
35 |
+
outputs = model(**inputs)
|
36 |
+
logits = outputs.logits
|
37 |
+
predicted_class = logits.argmax(dim=-1).item()
|
38 |
+
|
39 |
+
# Display the prediction
|
40 |
+
if predicted_class == 0:
|
41 |
+
st.write("Prediction: Negative")
|
42 |
+
else:
|
43 |
+
st.write("Prediction: Positive")
|