done with functionality

app.py

@@ -0,0 +1,40 @@
+import gradio as gr
+import pandas as pd
+from io import BytesIO
+from transformers import AutoTokenizer, AutoModelForTableQuestionAnswering, TableQuestionAnsweringPipeline
+
+# Load the tokenizer and model directly
+tokenizer = AutoTokenizer.from_pretrained("google/tapas-large-finetuned-wikisql-supervised")
+model = AutoModelForTableQuestionAnswering.from_pretrained("google/tapas-large-finetuned-wikisql-supervised")
+
+# Initialize the TableQuestionAnsweringPipeline manually
+pipe = TableQuestionAnsweringPipeline(model=model, tokenizer=tokenizer)
+
+def answer_question(uploaded_file, question):
+    # Convert the binary stream to a file-like object
+    file_like = BytesIO(uploaded_file)
+    
+    # Read the uploaded file directly into a DataFrame
+    df = pd.read_csv(file_like)
+
+    # Convert all DataFrame elements to string, as TAPAS expects string inputs
+    df = df.astype(str)
+    
+    # Use the pipeline to answer the question based on the table
+    result = pipe({"table": df, "query": question})
+    
+    # Format the answer before returning it
+    answer = result['answer']
+    return answer
+
+# Define the Gradio app interface
+iface = gr.Interface(
+    fn=answer_question,
+    inputs=[gr.File(label="Upload CSV File", type="binary"), gr.Textbox(lines=2, placeholder="Ask a question...")],
+    outputs=gr.Text(),
+    title="Table-based Question Answering",
+    description="Upload a CSV file and ask a question related to the data in the file."
+)
+
+# Run the app
+iface.launch()