Update app.py

app.py

@@ -9,6 +9,7 @@ from langchain_huggingface import HuggingFaceEndpoint, HuggingFaceEmbeddings
 from langchain.chains import ConversationalRetrievalChain
 from langchain.memory import ConversationBufferMemory
 from langchain.prompts import PromptTemplate
+from langchain_core.documents import Document
 from pptx import Presentation
 from io import BytesIO
 import shutil
@@ -38,7 +39,7 @@ chat_history = []
 memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
 PERSIST_DIRECTORY = tempfile.mkdtemp()  # Use temporary directory for ChromaDB
 
-# Custom prompt template
+# Custom prompt templates
 CONDENSE_QUESTION_PROMPT = PromptTemplate.from_template(
     """Given the following conversation and a follow-up question, rephrase the follow-up question to be a standalone question, incorporating relevant context from the conversation.
 
@@ -77,7 +78,10 @@ class PPTXLoader:
                         if hasattr(shape, "text") and shape.text:
                             text += shape.text + "\n"
                     if text.strip():
-                        docs.append({"page_content": text, "metadata": {"source": self.file_path, "slide": slide_num}})
+                        docs.append(Document(
+                            page_content=text,
+                            metadata={"source": self.file_path, "slide": slide_num}
+                        ))
         except Exception as e:
             logger.error(f"Error loading PPTX {self.file_path}: {str(e)}")
             return []
@@ -101,7 +105,7 @@ def load_documents(files):
                 documents.extend(loader.load())
             elif file_path.endswith(".pptx"):
                 loader = PPTXLoader(file_path)
-                documents.extend([{"page_content": doc["page_content"], "metadata": doc["metadata"]} for doc in loader.load()])
+                documents.extend(loader.load())
         except Exception as e:
             logger.error(f"Error loading file {file_path}: {str(e)}")
             continue