Update app.py

app.py

@@ -2,21 +2,44 @@ import gradio as gr
 import torch
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 
-# Load Pretrained Model & Tokenizer (XLM-Roberta for multilingual text classification)
+# Load Pretrained Model & Tokenizer (Ensure this is a fine-tuned model)
 MODEL_NAME = "xlm-roberta-base"
-model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME, num_labels=5)
+model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME, num_labels=5)  # Adjust num_labels as per training
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
 
+# Define Label Mapping (Modify based on your dataset)
+LABEL_MAPPING = {
+    0: "Contract",
+    1: "Invoice",
+    2: "Financial Report",
+    3: "Legal Notice",
+    4: "Marketing Material"
+}
+
 # Classification Function
 def classify_text(text):
     inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True, max_length=512)
+    
     with torch.no_grad():
         outputs = model(**inputs)
-    label = torch.argmax(outputs.logits, dim=1).item()
-    return f"Predicted Category: {label}"
+    
+    # Convert logits to probabilities
+    probs = torch.nn.functional.softmax(outputs.logits, dim=1)
+    
+    # Get predicted label index
+    label_idx = torch.argmax(probs, dim=1).item()
+    
+    # Retrieve category name
+    category = LABEL_MAPPING.get(label_idx, "Unknown")
+    
+    return f"Predicted Category: {category} (Confidence: {probs[0][label_idx]:.2f})"
 
 # Gradio UI
-demo = gr.Interface(fn=classify_text, inputs=gr.Textbox(lines=2, placeholder="Enter business document text..."), 
-                    outputs="text", title="Multilingual Business Document Classifier")
+demo = gr.Interface(
+    fn=classify_text, 
+    inputs=gr.Textbox(lines=4, placeholder="Enter business document text..."), 
+    outputs="text", 
+    title="Multilingual Business Document Classifier"
+)
 
-demo.launch()
+demo.launch()