Create app.py

app.py

@@ -0,0 +1,50 @@
+import streamlit as st
+import tensorflow as tf
+from transformers import BertTokenizer, TFBertForSequenceClassification
+from tensorflow.keras.preprocessing.sequence import pad_sequences
+
+# Load the BERT tokenizer and model
+tokenizer = BertTokenizer.from_pretrained('config.json')  # Path to folder containing config.json
+model = TFBertForSequenceClassification.from_pretrained('tf_model.h5', from_pt=True)  # Path to folder containing tf_model.h5
+
+def predict(text):
+    # Encode the text using the BERT tokenizer
+    input_ids = tokenizer.encode(text, add_special_tokens=True, max_length=128, truncation=True)
+    input_ids = pad_sequences([input_ids], maxlen=128, truncating='post', padding='post')
+    
+    # Convert to tensors
+    input_ids = tf.convert_to_tensor(input_ids)
+    
+    # Get predictions
+    logits = model(input_ids)[0]
+    
+    # Apply softmax to calculate probabilities
+    probabilities = tf.nn.softmax(logits, axis=1).numpy()[0]
+    
+    return probabilities
+
+# Streamlit UI
+st.title("Stress Categorization with BERT")
+st.write("Enter the text to analyze for stress levels:")
+
+# Text input
+user_input = st.text_area("Text", height=150)
+
+if st.button("Predict"):
+    # Make prediction
+    probabilities = predict(user_input)
+    
+    # Display probabilities
+    st.write("Probabilities:")
+    st.write(f"Stressed: {probabilities[1]:.4f}")
+    st.write(f"Not Stressed: {probabilities[0]:.4f}")
+    
+    # Display the most likely class
+    if probabilities[0] > probabilities[1]:
+        st.success("Prediction: Not Stressed")
+    else:
+        st.error("Prediction: Stressed")
+
+    # Assuming you have an accuracy metric available (replace with actual accuracy if available)
+    accuracy = 0.95  # Example accuracy
+    st.write(f"Model Accuracy: {accuracy * 100:.2f}%")