Upload app.py

app.py

@@ -0,0 +1,86 @@
+import streamlit as st
+import torch
+from transformers import BertTokenizer, BertForSequenceClassification
+
+# Set the title of the app
+st.title('Sentiment Analysis with BERT')
+
+# Load the tokenizer and model
+model_dir = './saved_model/'  # Update this path if your model is saved elsewhere
+
+@st.cache_resource
+def load_model():
+    tokenizer = BertTokenizer.from_pretrained(model_dir)
+    model = BertForSequenceClassification.from_pretrained(model_dir)
+    model.eval()
+    # Move the model to the appropriate device
+    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+    model.to(device)
+    return tokenizer, model, device
+
+tokenizer, model, device = load_model()
+
+# Function to perform sentiment analysis
+def predict_sentiment(texts):
+    # Tokenize and encode the texts
+    inputs = tokenizer(
+        texts,
+        padding=True,
+        truncation=True,
+        max_length=128,
+        return_tensors='pt'
+    )
+
+    # Move inputs to the same device as the model
+    inputs = {key: val.to(device) for key, val in inputs.items()}
+
+    # Perform inference
+    with torch.no_grad():
+        outputs = model(**inputs)
+        logits = outputs.logits
+        probabilities = torch.softmax(logits, dim=-1)
+        predicted_classes = torch.argmax(probabilities, dim=-1)
+        confidences = torch.max(probabilities, dim=-1).values
+
+    # Map predictions to labels
+    label_map = {0: 'Negative', 1: 'Positive'}
+    predicted_labels = [label_map[pred.item()] for pred in predicted_classes]
+    confidences = confidences.cpu().numpy()
+
+    return predicted_labels, confidences
+
+# Example texts (if you want to include them in the app)
+example_texts = [
+    "The bouquet was absolutely stunning! Fresh flowers and beautiful arrangement.",
+    "Delivery was late and the flowers were wilted. Very disappointed.",
+    "Great service and reasonable prices. Will order again!",
+    "The flowers didn't look like the picture at all. False advertising.",
+    "Prompt delivery and the roses lasted for weeks. Excellent quality!",
+    "Customer service was unhelpful when I had an issue with my order.",
+    "The arrangement was okay, but overpriced for what I received.",
+    "Impressed with the variety of flowers available. Something for every occasion!",
+    "The online ordering process was confusing and frustrating.",
+    "Received compliments on the centerpiece all night. Thank you for making our event special!",
+]
+
+# Add a select box for example texts
+selected_example = st.selectbox("Or select an example text:", [""] + example_texts)
+
+# Text input from the user
+if selected_example != "":
+    user_input = selected_example
+else:
+    user_input = st.text_area("Enter text for sentiment analysis:", "")
+
+if st.button("Analyze"):
+    if user_input.strip() == "":
+        st.warning("Please enter text to analyze.")
+    else:
+        with st.spinner('Analyzing...'):
+            # Since the function expects a list of texts, wrap the user_input in a list
+            labels, confidences = predict_sentiment([user_input])
+            label = labels[0]
+            confidence = confidences[0]
+            st.success(f"Predicted Sentiment: **{label}**")
+            st.info(f"Confidence: {confidence * 100:.2f}%")
+            st.write(f"Text: {user_input}")