Update app.py

app.py

@@ -1,44 +1,32 @@
+# imports
 import streamlit as st
-from streamlit_chat import message
-import os, tempfile, sys
-from io import BytesIO
-from io import StringIO
+import os, tempfile
 import pandas as pd
-from langchain.agents import create_pandas_dataframe_agent
-from langchain.llms.openai import OpenAI
-from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain_openai import ChatOpenAI
+from langchain_community.vectorstores import FAISS
+from langchain_openai import OpenAIEmbeddings
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import CSVLoader
+from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
+from langchain_core.runnables.history import RunnableWithMessageHistory
+from langchain_community.chat_message_histories import ChatMessageHistory
+from langchain_core.chat_history import BaseChatMessageHistory
 from langchain.chains.summarize import load_summarize_chain
-from langchain.document_loaders.csv_loader import CSVLoader
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain.text_splitter import CharacterTextSplitter
-from langchain.chains.mapreduce import MapReduceChain
-from langchain.docstore.document import Document
-from langchain.vectorstores import FAISS
-from langchain.chat_models import ChatOpenAI
-from langchain.chains import ConversationalRetrievalChain
-from langchain.chains import RetrievalQA
-from langchain.memory import ConversationBufferMemory
-from langchain.chains.conversational_retrieval.prompts import CONDENSE_QUESTION_PROMPT
-from langchain.chains.question_answering import load_qa_chain
-from langchain.prompts.prompt import PromptTemplate
-from langchain import LLMChain
-
+from langchain_core.prompts import MessagesPlaceholder
+from langchain_experimental.agents import create_pandas_dataframe_agent
+import asyncio
 
 st.set_page_config(page_title="CSV AI", layout="wide")
-
+os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
 
 def home_page():
  st.write("""Select any one feature from above sliderbox: \n
  1. Chat with CSV \n
  2. Summarize CSV \n
  3. Analyze CSV """)
 
-def chat(temperature, model_name):
- st.write("# Talk to CSV")
- # Add functionality for Page 1
- reset = st.sidebar.button("Reset Chat")
- uploaded_file = st.sidebar.file_uploader("Upload your CSV here 👇:", type="csv")
-
+@st.cache_resource()
+def retriever_func(uploaded_file):
  if uploaded_file :
  with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
  tmp_file.write(uploaded_file.getvalue())
@@ -50,60 +38,92 @@ def chat(temperature, model_name):
  loader = CSVLoader(file_path=tmp_file_path, encoding="cp1252")
  data = loader.load()
 
- embeddings = OpenAIEmbeddings()
- vectors = FAISS.from_documents(data, embeddings)
- llm = ChatOpenAI(temperature=temperature, model_name=model_name) # 'gpt-3.5-turbo',
- qa = RetrievalQA.from_chain_type(llm=llm,
- chain_type="stuff", 
- retriever=vectors.as_retriever(), 
- verbose=True)
- #chain = ConversationalRetrievalChain.from_llm(llm = ChatOpenAI(temperature=temperature, model_name=model_name), retriever=vectors.as_retriever())
+ text_splitter = RecursiveCharacterTextSplitter(
+ chunk_size=1000, 
+ chunk_overlap=200, 
+ add_start_index=True
+ )
+ all_splits = text_splitter.split_documents(data)
 
- def conversational_chat(query):
 
-# result = chain({"question": query, "chat_history": st.session_state['history']})
-# st.session_state['history'].append((query, result["answer"]))
- result = qa.run(query) #chain({"question": query, "chat_history": st.session_state['history']})
- st.session_state['history'].append((query, result))#["answer"]))
+ vectorstore = FAISS.from_documents(documents=all_splits, embedding=OpenAIEmbeddings())
+ retriever = vectorstore.as_retriever(search_type="similarity", search_kwargs={"k": 6})
+ if not uploaded_file:
+ st.info("Please upload CSV documents to continue.")
+ st.stop()
+ return retriever, vectorstore
+
+def chat(temperature, model_name):
+ st.write("# Talk to CSV")
+ # Add functionality for Page 1
+ reset = st.sidebar.button("Reset Chat")
+ uploaded_file = st.sidebar.file_uploader("Upload your CSV here 👇:", type="csv")
+ retriever, vectorstore = retriever_func(uploaded_file)
+ llm = ChatOpenAI(model_name=model_name, temperature=temperature, streaming=True)
 
- return result#["answer"]
+ def format_docs(docs):
+ return "\n\n".join(doc.page_content for doc in docs)
 
- if 'history' not in st.session_state:
- st.session_state['history'] = []
-
- if 'generatedd' not in st.session_state:
- st.session_state['generatedd'] = ["Hello ! Ask me anything about " + uploaded_file.name + " 🤗"]
+ if "messages" not in st.session_state:
+ st.session_state["messages"] = [{"role": "assistant", "content": "How can I help you?"}]
+
+ store = {}
+
+ prompt = ChatPromptTemplate.from_messages(
+ [
+ (
+ "system",
+ """Use the following pieces of context to answer the question at the end.
+ If you don't know the answer, just say that you don't know, don't try to make up an answer. Context: {context}""",
+ ),
+ MessagesPlaceholder(variable_name="history"),
+ ("human", "{input}"),
+ ]
+ )
+ runnable = prompt | llm
+
+ def get_session_history(session_id: str) -> BaseChatMessageHistory:
+ if session_id not in store:
+ store[session_id] = ChatMessageHistory()
+ return store[session_id]
 
- if 'pastt' not in st.session_state:
- st.session_state['pastt'] = ["Hey ! 👋"]
-
- #container for the chat history
- response_container = st.container()
- #container for the user's text input
- container = st.container()
 
- with container:
- with st.form(key='my_form', clear_on_submit=True):
-
- user_input = st.text_input("Query:", placeholder="Talk about your csv data here (:", key='input')
- submit_button = st.form_submit_button(label='Send')
-
- if submit_button and user_input:
- output = conversational_chat(user_input)
-
- st.session_state['pastt'].append(user_input)
- st.session_state['generatedd'].append(output)
+ with_message_history = RunnableWithMessageHistory(
+ runnable,
+ get_session_history,
+ input_messages_key="input",
+ history_messages_key="history",
+ )
 
- if st.session_state['generatedd']:
- with response_container:
- for i in range(len(st.session_state['generatedd'])):
- message(st.session_state["pastt"][i], is_user=True, key=str(i) + '_user', avatar_style="fun-emoji")
- message(st.session_state["generatedd"][i], key=str(i), avatar_style="bottts")
+ for msg in st.session_state.messages:
+ st.chat_message(msg["role"]).write(msg["content"])
+ async def chat_message():
+ if prompt := st.chat_input():
+ if not user_api_key: 
+ st.info("Please add your OpenAI API key to continue.")
+ st.stop()
+ st.session_state.messages.append({"role": "user", "content": prompt})
+ st.chat_message("user").write(prompt)
+ contextt = vectorstore.similarity_search(prompt, k=6)
+ context = "\n\n".join(doc.page_content for doc in contextt)
+ #msg = 
+ with st.chat_message("assistant"):
+ message_placeholder = st.empty()
+ text_chunk = ""
+ async for chunk in with_message_history.astream(
+ {"context": context, "input": prompt},
+ config={"configurable": {"session_id": "abc123"}},
+ ):
+ text_chunk += chunk.content
+ message_placeholder.markdown(text_chunk)
+ #st.chat_message("assistant").write(text_chunk)
+ st.session_state.messages.append({"role": "assistant", "content": text_chunk})
  if reset:
- st.session_state["pastt"] = []
- st.session_state["generatedd"] = []
+ st.session_state["messages"] = []
+ asyncio.run(chat_message())
+
 
-def summary(model_name, temperature, top_p, freq_penalty):
+def summary(model_name, temperature, top_p):
  st.write("# Summary of CSV")
  st.write("Upload your document here:")
  uploaded_file = st.file_uploader("Upload source document", type="csv", label_visibility="collapsed")
@@ -112,7 +132,7 @@ def summary(model_name, temperature, top_p, freq_penalty):
  tmp_file.write(uploaded_file.getvalue())
  tmp_file_path = tmp_file.name
  # encoding = cp1252
- text_splitter = RecursiveCharacterTextSplitter(chunk_size = 500, chunk_overlap=0)
+ text_splitter = RecursiveCharacterTextSplitter(chunk_size = 1024, chunk_overlap=100)
  try:
  loader = CSVLoader(file_path=tmp_file_path, encoding="cp1252")
  #loader = UnstructuredFileLoader(tmp_file_path)
@@ -128,12 +148,18 @@ def summary(model_name, temperature, top_p, freq_penalty):
  gen_sum = st.button("Generate Summary")
  if gen_sum:
  # Initialize the OpenAI module, load and run the summarize chain
- llm = OpenAI(model_name=model_name, temperature=temperature)
- chain = load_summarize_chain(llm, chain_type="stuff")
- #search = docsearch.similarity_search(" ")
- summary = chain.run(input_documents=texts[:50])
+ llm = ChatOpenAI(model_name=model_name, temperature=temperature)
+ chain = load_summarize_chain(
+  llm=llm,
+  chain_type="map_reduce",
 
- st.success(summary)
+ return_intermediate_steps=True,
+ input_key="input_documents",
+ output_key="output_text",
+ )
+ result = chain({"input_documents": texts}, return_only_outputs=True)
+
+ st.success(result["output_text"])
 
 
 def analyze(temperature, model_name):
@@ -148,49 +174,26 @@ def analyze(temperature, model_name):
  tmp_file.write(uploaded_file.getvalue())
  tmp_file_path = tmp_file.name
  df = pd.read_csv(tmp_file_path)
+ llm = ChatOpenAI(model=model_name, temperature=temperature)
+ agent = create_pandas_dataframe_agent(llm, df, agent_type="openai-tools", verbose=True)
 
- def agent_chat(query):
-
- # Create and run the CSV agent with the user's query
- try:
- agent = create_pandas_dataframe_agent(ChatOpenAI(temperature=temperature, model_name=model_name), df, verbose=True, max_iterations=4)
- result = agent.run(query)
- except:
- result = "Try asking quantitative questions about structure of csv data!"
- return result
-
-
- if 'generated' not in st.session_state:
- st.session_state['generated'] = ["Hello ! Ask me anything about Document 🤗"]
-
- if 'past' not in st.session_state:
- st.session_state['past'] = ["Hey ! 👋"]
+ if "messages" not in st.session_state:
+ st.session_state["messages"] = [{"role": "assistant", "content": "How can I help you?"}]
 
- #container for the chat history
- response_container = st.container()
- #container for the user's text input
- container = st.container()
-
- with container:
- with st.form(key='my_form', clear_on_submit=True):
-
- user_input = st.text_input("Use CSV agent for precise information about the structure of your csv file:", placeholder="e-g : how many rows in my file ?:", key='input')
- submit_button = st.form_submit_button(label='Send')
-
- if submit_button and user_input:
- output = agent_chat(user_input)
-
- st.session_state['past'].append(user_input)
- st.session_state['generated'].append(output)
-
- if st.session_state['generated']:
- with response_container:
- for i in range(len(st.session_state['generated'])):
- message(st.session_state["past"][i], is_user=True, key=str(i) + '_user', avatar_style="big-smile")
- message(st.session_state["generated"][i], key=str(i), avatar_style="thumbs")
+ for msg in st.session_state.messages:
+ st.chat_message(msg["role"]).write(msg["content"])
+
+ if prompt := st.chat_input(placeholder="What are the names of the columns?"):
+ if not user_api_key: 
+ st.info("Please add your OpenAI API key to continue.")
+ st.stop()
+ st.session_state.messages.append({"role": "user", "content": prompt})
+ st.chat_message("user").write(prompt)
+ msg = agent.invoke({"input": prompt, "chat_history": st.session_state.messages})
+ st.session_state.messages.append({"role": "assistant", "content": msg["output"]})
+ st.chat_message("assistant").write(msg["output"])
  if reset:
- st.session_state["past"] = []
- st.session_state["generated"] = []
+ st.session_state["messages"] = []
 
 
 # Main App
@@ -211,8 +214,10 @@ def main():
  """,
  unsafe_allow_html=True,
  )
-
-
+ global user_api_key
+ # #
+ # st.sidebar.write("---Made with ❤️---")
+ # st.sidebar.write("---")
  if os.path.exists(".env") and os.environ.get("OPENAI_API_KEY") is not None:
  user_api_key = os.environ["OPENAI_API_KEY"]
  st.success("API key loaded from .env", icon="🚀")
@@ -228,15 +233,15 @@ def main():
 
 
  # Execute the home page function
- MODEL_OPTIONS = ["gpt-3.5-turbo", "gpt-4", "gpt-4-32k"]
+ MODEL_OPTIONS = ["gpt-3.5-turbo", "gpt-4", "gpt-4-32k","gpt-3.5-turbo-16k","gpt-4-1106-preview"]
  max_tokens = {"gpt-4":7000, "gpt-4-32k":31000, "gpt-3.5-turbo":3000}
  TEMPERATURE_MIN_VALUE = 0.0
  TEMPERATURE_MAX_VALUE = 1.0
  TEMPERATURE_DEFAULT_VALUE = 0.9
  TEMPERATURE_STEP = 0.01
  model_name = st.sidebar.selectbox(label="Model", options=MODEL_OPTIONS)
  top_p = st.sidebar.slider("Top_P", 0.0, 1.0, 1.0, 0.1)
- freq_penalty = st.sidebar.slider("Frequency Penalty", 0.0, 2.0, 0.0, 0.1)
+ # freq_penalty = st.sidebar.slider("Frequency Penalty", 0.0, 2.0, 0.0, 0.1)
  temperature = st.sidebar.slider(
  label="Temperature",
  min_value=TEMPERATURE_MIN_VALUE,
@@ -251,7 +256,7 @@ def main():
  "Summarize CSV",
  "Analyze CSV",
  ]
-
+ 
  #st.subheader("Select any generator👇")
  # Create a selectbox with the function names as options
  selected_function = st.selectbox("Select a functionality", functions)
@@ -260,15 +265,14 @@ def main():
  elif selected_function == "Chat with CSV":
  chat(temperature=temperature, model_name=model_name)
  elif selected_function == "Summarize CSV":
- summary(model_name=model_name, temperature=temperature, top_p=top_p, freq_penalty=freq_penalty)
+ summary(model_name=model_name, temperature=temperature, top_p=top_p)
  elif selected_function == "Analyze CSV":
  analyze(temperature=temperature, model_name=model_name)
  else:
  st.warning("You haven't selected any AI Functionality!!")
 
 
- st.write("---")
- st.write("Made with ❤️")
+
 
 if __name__ == "__main__":
  main()