Update app.py

app.py

@@ -1,142 +1,14 @@
-from langchain.embeddings.openai import OpenAIEmbeddings
-from langchain.text_splitter import RecursiveCharacterTextSplitter, CharacterTextSplitter
-from langchain.vectorstores import Chroma
-from langchain.chains import RetrievalQAWithSourcesChain
-from langchain.memory import ConversationBufferWindowMemory
-from langchain.chains import ConversationalRetrievalChain
-from langchain.chat_models import ChatOpenAI
-from langchain.prompts.chat import (
-    ChatPromptTemplate,
-    SystemMessagePromptTemplate,
-    HumanMessagePromptTemplate,
-)
-from langchain.document_loaders import PyPDFLoader
-import os
-import chainlit as cl
-from langchain.prompts import PromptTemplate
-
-text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
 
-system_template = """Use the following pieces of context to answer the users question.
-If you don't know the answer, just say that you don't know, don't try to make up an answer.
-ALWAYS return a "SOURCES" part in your answer.
-The "SOURCES" part should be a reference to the source of the document from which you got your answer.
-Example of your response should be:
-```
-The answer is foo
-SOURCES: xyz
-```
-Begin!
-----------------
-{summaries}"""
-messages = [
-    SystemMessagePromptTemplate.from_template(system_template),
-    HumanMessagePromptTemplate.from_template("{question}"),
-]
-prompt = ChatPromptTemplate.from_messages(messages)
-chain_type_kwargs = {"prompt": prompt}
+import chainlit as cl
 
 @cl.on_chat_start
-async def start():
-    await cl.Avatar(
-        name="ChatPDF",
-        url="https://avatars.githubusercontent.com/u/128686189?s=400&u=a1d1553023f8ea0921fba0debbe92a8c5f840dd9&v=4",
-        # path = r'assets/ChatPDFAvatar.jpg'
-    ).send()
-
-
-@cl.langchain_factory(use_async=True)
-async def init():
-    files = None
-
-    # Wait for the user to upload a file
-    while files == None:
-        files = await cl.AskFileMessage(
-            content="Hey, Welcome to ChatPDF!\n\nChatPDF is a smart, user-friendly tool that integrates state-of-the-art AI models with text extraction and embedding capabilities to create a unique, conversational interaction with your PDF documents.\n\nSimply upload your PDF, ask your questions, and ChatPDF will deliver the most relevant answers directly from your document.\n\nPlease upload a PDF file to begin!",max_size_mb=100, accept=["application/pdf"]
-        ).send()
-
-    file = files[0]
-
-    msg = cl.Message(content=f'''Processing "{file.name}"...''')
-    await msg.send()
-
-    #
-    
-    with open(os.path.join(file.name), "wb") as f:
-        f.write(file.content)
-
-    print(file.name)
+async def on_chat_start():
     
-    loader = PyPDFLoader(file.name)
-    pages = loader.load_and_split()
-
-    # add page split info
-    # Initialize a dictionary to keep track of duplicate page numbers
-    page_counts = {}
-
-    for document in pages:
-        page_number = document.metadata['page']
-
-        # If this is the first occurrence of this page number, initialize its count to 1
-        # Otherwise, increment the count for this page number
-        page_counts[page_number] = page_counts.get(page_number, 0) + 1
+    await cl.Message(content=f'Hello, How can I help you today?').send()
 
-        # Create the page split info string
-        page_split_info = f"Page-{page_number+1}.{page_counts[page_number]}"
+@cl.on_message
+async def on_message(message: cl.Message):
 
-        # Add the page split info to the document's metadata
-        document.metadata['page_split_info'] = page_split_info
-
-    
-
-    # Create a Chroma vector store
-    embeddings = OpenAIEmbeddings()
-    docsearch = await cl.make_async(Chroma.from_documents)(
-        pages, embeddings
-    )
-
-    # define memory
-    memory = ConversationBufferWindowMemory(
-        k=5,
-        memory_key='chat_history', 
-        return_messages=True, 
-        output_key='answer'
-        )
-    
-    # Create a chain that uses the Chroma vector store
-    chain = ConversationalRetrievalChain.from_llm(
-        ChatOpenAI(temperature=0, model="gpt-3.5-turbo-16k", streaming=True),
-        chain_type="stuff",
-        retriever=docsearch.as_retriever(search_kwargs={'k':10}),
-        memory=memory,
-        return_source_documents=True,
-    )
-
-    # Save the metadata and texts in the user session
-    # cl.user_session.set("metadatas", metadatas)
-    cl.user_session.set("texts", pages)
-
-    # Let the user know that the system is ready
-    await msg.update(content=f''' "{file.name}" processed. You can now ask questions!''')
-
-
-    return chain
-
-
-@cl.langchain_postprocess
-async def process_response(res):
-    answer = res["answer"]
-    source_documents = res['source_documents']
-    content = [source_documents[i].page_content for i in range(len(source_documents))]
-    name = [source_documents[i].metadata['page_split_info'] for i in range(len(source_documents))]
-    source_elements = [
-        cl.Text(content=content[i], name=name[i]) for i in range(len(source_documents))
-        ]
-    
-    if source_documents:
-        answer += f"\n\nSources: {', '.join([source_documents[i].metadata['page_split_info'] for i in range(len(source_documents))])}"
-    else:
-        answer += "\n\nNo sources found"
-    
-    await cl.Message(content=answer, elements=source_elements).send()
-    # await cl.Message(content=answer).send()
+    await cl.Message(
+        content=f"Received: {message.content}",
+    ).send()