simple stuffing

app.py

@@ -36,10 +36,6 @@ def pdf_changes(pdf_doc):
         loader = OnlinePDFLoader(pdf_doc.name)
         documents = loader.load()
 
-        # Split loaded documents into chunks
-        text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
-        texts = text_splitter.split_documents(documents)
-
         # Define the prompt for summarization
         prompt_template = """Write a concise summary of the following:
         "{text}"
@@ -55,23 +51,13 @@ def pdf_changes(pdf_doc):
             llm_chain=llm_chain, document_variable_name="text"
         )
 
-        # Initialize summary variable
-        full_summary = ""
-
-        # Iterate through text chunks to summarize
-        for i in range(0, len(texts), 2):
-            chunk = " ".join([doc.page_content for doc in texts[i:i + 2]])
-
-            # Generate summary using StuffDocumentsChain
-            chunk_summary = stuff_chain.run([chunk])
-            
-            # Add chunk summary to full summary
-            full_summary += f"Summary of pages {i+1}-{i+3}:\n{chunk_summary}\n"
+        # Generate summary using StuffDocumentsChain
+        full_summary = stuff_chain.run(documents)
 
         # Other existing logic for Chroma, embeddings, and retrieval
         embeddings = OpenAIEmbeddings()
         global db
-        db = Chroma.from_documents(texts, embeddings)
+        db = Chroma.from_documents(documents, embeddings)
 
         retriever = db.as_retriever()
         global qa
@@ -88,7 +74,6 @@ def pdf_changes(pdf_doc):
 
 
 
-
 def clear_data():
     global qa, db
     qa = None