Create app.py

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+from transformers import T5ForConditionalGeneration, T5Tokenizer
+from textwrap import fill
+
+# Load fine-tuned model and tokenizer
+last_checkpoint = "Jyotiyadav/InsuranceModel1.0"
+finetuned_model = T5ForConditionalGeneration.from_pretrained(last_checkpoint)
+tokenizer = T5Tokenizer.from_pretrained(last_checkpoint)
+
+# Define inference function
+def answer_question(question):
+    # Format input
+    inputs = ["Please answer this question: " + question]
+    inputs = tokenizer(inputs, return_tensors="pt")
+    
+    # Generate answer
+    outputs = finetuned_model.generate(**inputs)
+    answer = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    
+    # Wrap answer for better display
+    return fill(answer, width=80)
+
+# Create Gradio interface
+iface = gr.Interface(
+    fn=answer_question,
+    inputs="text",
+    outputs="text",
+    title="Question Answering with T5 Model",
+    description="Enter your question to get the answer.",
+    examples=[
+        ["For a Male customer with an annual income of $850000, who bought a Pale White Mitsubishi Diamante (Overhead Camshaft engine) from Classic Chevy in Riga on 2022-Jan-2, priced at $12000, what was the claim amount?"],
+        ["For a Male customer with an annual income of $13500, who bought a Pale White Chrysler Sebring Coupe (Overhead Camshaft engine) from Suburban Ford in Ventspils on 2022-Jan-3, priced at $26000, what was the claim amount?"],
+        ["For a Male customer with an annual income of $13500, who bought a Black Lexus LS400 (Double\u00c3\u201a\u00c2\u00a0Overhead Camshaft engine) from Saab-Belle Dodge in Liepaja on 2022-Jan-12, priced at $39000, what was the claim amount?"]
+    ]
+)
+
+# Launch Gradio interface
+iface.launch()