Update app.py

app.py

@@ -1,4 +1,4 @@
-import spaces  # MUST BE AT LINE 1 BEFORE ANY PYTORCH / LLAMA-INDEX IMPORTS
+import spaces
 
 import os
 import subprocess
@@ -10,7 +10,6 @@ from llama_index.vector_stores.qdrant import QdrantVectorStore
 from qdrant_client import QdrantClient
 from qdrant_client.models import Distance, VectorParams
 
-# 1. Initialize Global Settings (forcing embeddings to CPU)
 Settings.llm = Groq(model="llama3-8b-8192", api_key=os.environ.get("GROQ_API_KEY"))
 Settings.embed_model = HuggingFaceEmbedding(
     model_name="BAAI/bge-large-en-v1.5",
@@ -19,7 +18,6 @@ Settings.embed_model = HuggingFaceEmbedding(
 Settings.chunk_size = 512
 Settings.chunk_overlap = 50
 
-# 2. Setup In-Memory Qdrant and Index
 client = QdrantClient(location=":memory:")
 client.create_collection(
     collection_name="tech_docs",
@@ -28,7 +26,6 @@ client.create_collection(
 vector_store = QdrantVectorStore(client=client, collection_name="tech_docs")
 index = VectorStoreIndex.from_vector_store(vector_store=vector_store)
 
-# 3. Ingest Default GitHub Repo on Startup (FastAPI Docs)
 TARGET_REPO = "https://github.com/tiangolo/fastapi.git"
 clone_path = "/tmp/fastapi_docs"
 
@@ -43,21 +40,18 @@ try:
         required_exts=[".md", ".mdx"], 
         recursive=True
     ).load_data()
-    # Ingest a subset to fit comfortably within free RAM constraints
+
     for doc in documents[:200]:
         index.insert(doc)
     print("GitHub repository indexed successfully!")
 except Exception as e:
     print(f"Error indexing GitHub repo: {e}")
 
-# Create global query engine
 query_engine = index.as_query_engine(streaming=False)
 
-# 4. Define Response Function for Gradio Chat Interface
 def chat_with_docs(message, history, file_obj):
     global index, query_engine
     
-    # If user uploaded a new PDF, ingest it dynamically into the active session
     if file_obj is not None:
         try:
             pdf_docs = SimpleDirectoryReader(input_files=[file_obj.name]).load_data()
@@ -67,11 +61,9 @@ def chat_with_docs(message, history, file_obj):
         except Exception as e:
             return f"Error processing uploaded PDF: {str(e)}"
             
-    # Query the RAG engine using Groq acceleration
     response = query_engine.query(message)
     return str(response)
 
-# 5. Build Gradio Chat UI Layout
 demo = gr.ChatInterface(
     fn=chat_with_docs,
     title="📚 TechDocs-Knowledge-Engine",