Update app.py

app.py

@@ -9,26 +9,21 @@ import torch
 import gradio as gr
 
 # Load dataset
-try:
-    dataset = pd.read_csv("dataset.csv")
-except FileNotFoundError:
-    raise FileNotFoundError("❌ dataset.csv not found. Please upload it alongside app.py.")
+dataset = pd.read_csv("dataset.csv")
 
-# Convert dataset to LangChain Documents
-def dataframe_to_documents(df, content_col="answer", metadata_cols=["question"]):
-    return [
-        Document(
-            page_content=str(row[content_col]),
-            metadata={col: str(row[col]) for col in metadata_cols}
-        )
-        for _, row in df.iterrows()
-    ]
+# Convert to LangChain Documents
+documents = [
+    Document(
+        page_content=str(row["answer"]),
+        metadata={"question": str(row["question"])}
+    )
+    for _, row in dataset.iterrows()
+]
 
-documents = dataframe_to_documents(dataset)
-
-# Initialize components
+# Setup embeddings
 embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
 
+# Load LLM
 model_name = "google/flan-t5-base"
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
@@ -36,44 +31,39 @@ pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer, max_le
                 device=0 if torch.cuda.is_available() else -1)
 llm = HuggingFacePipeline(pipeline=pipe)
 
-# Create vector store & retriever
+# Create vector store
 vector_store = FAISS.from_documents(documents, embeddings)
 retriever = vector_store.as_retriever(search_kwargs={"k": 3})
 
-# Create QA Chain
+# Create QA chain
 qa_chain = RetrievalQA.from_chain_type(
     llm=llm,
     retriever=retriever,
     return_source_documents=True
 )
 
-# Chatbot logic
+# Chatbot function
 def chatbot_interface(question: str) -> str:
-    if not question.strip():
-        return "❗ Please enter a question."
-
+    if not qa_chain:
+        return "Chatbot backend not initialized properly."
     try:
         response = qa_chain.invoke({"query": question})
-        answer = response.get("result", "❌ No answer found.")
+        answer = response.get("result", "No answer found.")
         sources = response.get("source_documents", [])
         source_texts = [doc.page_content for doc in sources]
-
-        # Format response
-        formatted_sources = "\n".join(f"- {src}" for src in source_texts) if source_texts else "No sources found."
-        return f"✅ **Answer:** {answer}\n\n📚 **Sources:**\n{formatted_sources}"
-
+        return f"Answer: {answer}\n\nSources:\n" + "\n".join(f"- {text}" for text in source_texts)
     except Exception as e:
-        return f"❌ Error: {e}"
+        return f"Error: {e}"
 
 # Gradio UI
 interface = gr.Interface(
     fn=chatbot_interface,
-    inputs=gr.Textbox(label="Ask a question", placeholder="e.g., What is Artificial Intelligence?"),
+    inputs=gr.Textbox(label="Enter your question"),
     outputs=gr.Textbox(label="Response"),
-    title="RAG Chatbot 🤖",
-    description="Ask me anything about AI, RAG, NLP, and more!",
+    title="RAG Chatbot",
+    description="Ask questions about AI, ChatBots, NLP, and more.",
     theme="default"
 )
 
-# Run the app
+# Launch the interface
 interface.launch()