Remove the Error

app.py

@@ -1,17 +1,19 @@
-import gradio as gr
 import os
 import tempfile
 from typing import List
+
+import gradio as gr
 import PyPDF2
-from sentence_transformers import SentenceTransformer
 import numpy as np
-from sklearn.metrics.pairwise import cosine_similarity
 
+from sentence_transformers import SentenceTransformer
+from sklearn.metrics.pairwise import cosine_similarity
 from groq import Groq
-embedder = SentenceTransformer("all-MiniLM-L6-v2")
-# -----------------------------
+
+# --------------------------------------------------
 # Configuration
-# -----------------------------
+# --------------------------------------------------
+
 GROQ_API_KEY = os.getenv("GROQ_API_KEY")
 
 if not GROQ_API_KEY:
@@ -22,73 +24,84 @@ if not GROQ_API_KEY:
 MODEL_NAME = "llama-3.1-8b-instant"
 
 client = Groq(api_key=GROQ_API_KEY)
+embedder = SentenceTransformer("all-MiniLM-L6-v2")
 
-
-# -----------------------------
+# --------------------------------------------------
 # PDF Processing
-# -----------------------------
+# --------------------------------------------------
+
 def extract_text_from_pdfs(files: List[tempfile.NamedTemporaryFile]):
     documents = []
+
     for file in files:
         reader = PyPDF2.PdfReader(file)
         for page_num, page in enumerate(reader.pages):
             text = page.extract_text()
-            if text:
+            if text and text.strip():
                 documents.append({
                     "text": text,
                     "source": f"{os.path.basename(file.name)} - page {page_num + 1}"
                 })
+
     return documents
 
-# -----------------------------
+# --------------------------------------------------
 # Chunking
-# -----------------------------
+# --------------------------------------------------
+
 def chunk_text(documents, chunk_size=500, overlap=50):
     chunks = []
+
     for doc in documents:
-        text = doc["text"]
-        words = text.split()
+        words = doc["text"].split()
         start = 0
+
         while start < len(words):
             chunk_words = words[start:start + chunk_size]
-            chunk_text = " ".join(chunk_words)
+            chunk = " ".join(chunk_words)
+
             chunks.append({
-                "text": chunk_text,
+                "text": chunk,
                 "source": doc["source"]
             })
+
             start += chunk_size - overlap
+
     return chunks
 
-# -----------------------------
+# --------------------------------------------------
 # Embeddings & Retrieval
-# -----------------------------
+# --------------------------------------------------
+
 def embed_chunks(chunks):
     texts = [c["text"] for c in chunks]
-    embeddings = embedder.encode(texts)
-    return embeddings
+    return embedder.encode(texts)
 
 def retrieve_relevant_chunks(query, chunks, embeddings, top_k=3):
     query_embedding = embedder.encode([query])
     similarities = cosine_similarity(query_embedding, embeddings)[0]
+
     top_indices = np.argsort(similarities)[-top_k:][::-1]
 
-    results = []
-    for idx in top_indices:
-        results.append(chunks[idx])
-    return results
+    return [chunks[i] for i in top_indices]
 
-# -----------------------------
+# --------------------------------------------------
 # LLM Call
-# -----------------------------
-def ask_llm(question, context, history):
-    messages = history.copy()
+# --------------------------------------------------
 
-    system_prompt = (
-        "You are a helpful assistant. Answer the question strictly using the provided context. "
-        "If the answer is not in the context, say so."
-    )
+def ask_llm(question, context, history):
+    messages = [
+        {
+            "role": "system",
+            "content": (
+                "You are a helpful assistant. Answer the question strictly using the provided context. "
+                "If the answer is not contained in the context, say so clearly."
+            )
+        }
+    ]
+
+    messages.extend(history)
 
-    messages.insert(0, {"role": "system", "content": system_prompt})
     messages.append({
         "role": "user",
         "content": f"Context:\n{context}\n\nQuestion:\n{question}"
@@ -102,14 +115,27 @@ def ask_llm(question, context, history):
 
     return response.choices[0].message.content
 
-# -----------------------------
-# Main Chat Logic
-# -----------------------------
+# --------------------------------------------------
+# Chat Logic (Gradio Compatible)
+# --------------------------------------------------
+
 def chat(files, question, chat_history):
     if not files:
-        return chat_history, "Please upload PDF files first."
+        chat_history.append({
+            "role": "assistant",
+            "content": "❌ Please upload at least one PDF file."
+        })
+        return chat_history
 
     documents = extract_text_from_pdfs(files)
+
+    if not documents:
+        chat_history.append({
+            "role": "assistant",
+            "content": "❌ Could not extract text from the uploaded PDF(s)."
+        })
+        return chat_history
+
     chunks = chunk_text(documents)
     embeddings = embed_chunks(chunks)
 
@@ -117,6 +143,7 @@ def chat(files, question, chat_history):
 
     context = ""
     sources = []
+
     for c in relevant_chunks:
         context += c["text"] + "\n\n"
         sources.append(c["source"])
@@ -124,19 +151,26 @@ def chat(files, question, chat_history):
     answer = ask_llm(question, context, chat_history)
 
     answer_with_sources = (
-        f"{answer}\n\n"
-        f"Sources:\n" + "\n".join(set(sources))
+        f"{answer}\n\n📄 Sources:\n" + "\n".join(set(sources))
     )
 
-    chat_history.append({"role": "user", "content": question})
-    chat_history.append({"role": "assistant", "content": answer_with_sources})
+    chat_history.append({
+        "role": "user",
+        "content": question
+    })
 
-    return chat_history, answer_with_sources
+    chat_history.append({
+        "role": "assistant",
+        "content": answer_with_sources
+    })
+
+    return chat_history
 
-# -----------------------------
+# --------------------------------------------------
 # Gradio UI
-# -----------------------------
-with gr.Blocks(title="Enhanced RAG Chatbot") as demo:
+# --------------------------------------------------
+
+with gr.Blocks(title="Enhanced RAG-Based Chatbot") as demo:
     gr.Markdown("## 📚 Enhanced RAG-Based Chatbot (PDF QA)")
     gr.Markdown("Upload multiple PDFs and ask questions based on their content.")
 
@@ -147,16 +181,15 @@ with gr.Blocks(title="Enhanced RAG Chatbot") as demo:
     )
 
     chatbot = gr.Chatbot(type="messages")
-    question = gr.Textbox(label="Ask a question")
+    question = gr.Textbox(label="Ask a question", placeholder="Type your question here...")
     state = gr.State([])
 
-    submit = gr.Button("Ask")
+    ask_button = gr.Button("Ask")
 
-    submit.click(
+    ask_button.click(
         fn=chat,
         inputs=[pdf_files, question, state],
-        outputs=[chatbot, chatbot]
+        outputs=chatbot
     )
 
-
 demo.launch()