Create app.py

app.py

@@ -0,0 +1,51 @@
+import streamlit as st
+import pandas as pd
+import numpy as np
+import torch
+import os
+from transformers import BertTokenizer, BertForSequenceClassification
+
+# Load pre-trained BERT model and tokenizer
+tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
+model = BertForSequenceClassification.from_pretrained('bert-base-uncased', num_labels=2)
+model.eval()
+
+# Define function to classify text
+def classify_text(text):
+    inputs = tokenizer(text, padding=True, truncation=True, return_tensors='pt')
+    outputs = model(**inputs)
+    probs = torch.softmax(outputs.logits, dim=1).detach().numpy()[0]
+    return probs[1]
+
+# Define function to save classification results to a persistent DataFrame
+def save_to_df(text, label):
+    data = {'text': [text], 'toxicity': [label]}
+    df = pd.DataFrame(data)
+    df.to_csv('results.csv', mode='a', header=not os.path.exists('results.csv'), index=False)
+
+# Load existing results from persistent DataFrame
+if os.path.exists('results.csv'):
+    results_df = pd.read_csv('results.csv')
+else:
+    results_df = pd.DataFrame(columns=['text', 'toxicity'])
+
+# Define Streamlit app
+def app():
+    st.title('Toxicity Classification App')
+
+    # Allow user to input text
+    text = st.text_input('Enter text to classify:')
+    if text:
+        # Classify text and display results
+        label = classify_text(text)
+        st.write(f'Toxic probability: {label:.2f}')
+        st.write('Classification: ', 'Toxic' if label >= 0.5 else 'Non-toxic')
+
+        # Allow user to save results
+        if st.button('Save results'):
+            save_to_df(text, label)
+            st.success('Results saved to DataFrame')
+
+    # Display existing results
+    st.subheader('Existing Results')
+    st.write(results_df)