Upload app.py

app.py

@@ -1,89 +1,152 @@
-import sqlite3
-import pandas as pd
-import openai
-import os
-import streamlit as st
-import datetime
- 
-# Set OpenAI API Key (Ensure it's set properly)
-openai.api_key = os.getenv("sk-NOBe-504FBda5dOQPesE8xKYgzmvBhG_Z_21UZGXtvT3BlbkFJF2vDvLgwzMMrIYgqAC6ezqMnupr9ZAelUCMH4XBP8A") or "sk-NOBe-504FBda5dOQPesE8xKYgzmvBhG_Z_21UZGXtvT3BlbkFJF2vDvLgwzMMrIYgqAC6ezqMnupr9ZAelUCMH4XBP8A"
- 
-# Create sample house data
-house_data = [
-    ("Alice", "1234567890", 15000, True, "Delhi", "2BHK", "Furnished"),
-    ("Bob", "9876543210", 45000, False, "Mumbai", "3BHK", "Semi Furnished"),
-    ("Charlie", "5556667777", 30000, True, "Pune", "4BHK", "Non Furnished"),
-    ("David", "4445556666", 25000, True, "Jaipur", "1BHK", "Furnished"),
-    ("Eve", "3332221111", 40000, False, "Ahmedabad", "3BHK", "Semi Furnished")
-]
- 
-# Create DataFrame and save to CSV
-df = pd.DataFrame(house_data, columns=["owner_name", "contact", "price", "for_sale", "location", "house_type", "house_details"])
-df.to_csv("houses.csv", index=False)
- 
-# Connect to SQLite Database
-conn = sqlite3.connect("houses.db")
-cursor = conn.cursor()
- 
-# Create Table if it doesn't exist
-cursor.execute('''
-    CREATE TABLE IF NOT EXISTS houses (
-        id INTEGER PRIMARY KEY AUTOINCREMENT,
-        owner_name TEXT,
-        contact TEXT,
-        price INTEGER,
-        for_sale BOOLEAN,
-        location TEXT,
-        house_type TEXT,
-        house_details TEXT
-    )
-''')
- 
-# Load data from CSV and insert into database (replace table content if exists)
-df = pd.read_csv("houses.csv")
-df.to_sql("houses", conn, if_exists="replace", index=False)
-conn.commit()
- 
-# Function to retrieve data from SQLite Database
-def retrieve_data():
-    cursor.execute("SELECT * FROM houses")
-    rows = cursor.fetchall()
-    # Assuming the table now includes the 'id' column, we include it in the DataFrame.
-    return pd.DataFrame(rows, columns=["owner_name", "contact", "price", "for_sale", "location", "house_type", "house_details"])
- 
-# Chatbot function that uses the database records in its prompt
-def chatbot(query):
-    df = retrieve_data()
-    relevant_data = df.to_string(index=False)
-    prompt = f"Given the following real estate records:\n{relevant_data}\n\nAnswer the user's query based on the provided data.\n\nUser Query: {query}\nResponse:"
-    
-    response = openai.ChatCompletion.create(
-        model="gpt-3.5-turbo",
-        messages=[
-            {"role": "system", "content": "You are an AI assistant that uses the provided database records to answer user queries."},
-            {"role": "user", "content": prompt}
-        ]
-    )
-    return response["choices"][0]["message"]["content"]
- 
-# Streamlit UI
-st.title("Real Estate Chatbot")
- 
-st.write("### Available Houses")
-st.dataframe(retrieve_data())
- 
-query = st.text_input("Ask about available houses:")
-if st.button("Ask Chatbot"):
-    if query:
-        response_text = chatbot(query)
-        # Get the current date and time
-        now = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
-        st.write("### User Query:")
-        st.write(query)
-        st.write(f"### Chatbot Response (Generated on {now}):")
-        st.write(response_text)
-    else:
-        st.warning("Please enter a query.")
- 
-conn.close()
- 
+from typing import Annotated, Sequence, TypedDict
+import functools
+import operator
+from bertopic import BERTopic
+from langchain.agents import AgentExecutor, create_openai_tools_agent
+from langchain_core.messages import BaseMessage, HumanMessage
+from langchain_openai import ChatOpenAI
+from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
+from langchain_core.output_parsers.openai_functions import JsonOutputFunctionsParser
+from langgraph.graph import END, StateGraph
+from langchain_community.tools.tavily_search import TavilySearchResults
+from langchain_experimental.tools import PythonREPLTool
+
+# Initialize tools
+tavily_tool = TavilySearchResults(max_results=5)
+python_repl_tool = PythonREPLTool()
+
+# Load BERTopic model
+topic_model = BERTopic.load("topic_model")
+
+# Function to create an agent
+def create_agent(llm: ChatOpenAI, tools: list, system_prompt: str):
+    prompt = ChatPromptTemplate.from_messages(
+        [
+            ("system", system_prompt),
+            MessagesPlaceholder(variable_name="messages"),
+            MessagesPlaceholder(variable_name="agent_scratchpad"),
+        ]
+    )
+    agent = create_openai_tools_agent(llm, tools, prompt)
+    executor = AgentExecutor(agent=agent, tools=tools)
+    return executor
+
+# Function to define an agent node
+def agent_node(state, agent, name):
+    result = agent.invoke(state)
+    return {"messages": [HumanMessage(content=result["output"], name=name)]}
+
+# Define the Viewer agent using the BERTopic model
+def viewer_agent(state):
+    review = state["messages"][-1].content
+    index = topic_model.transform([review])[0][0]
+    answer = topic_model.topic_labels_.get(index)
+    return {"messages": [HumanMessage(content=answer, name="Viewer")]}
+
+# Define AgentState
+class AgentState(TypedDict):
+    messages: Annotated[Sequence[BaseMessage], operator.add]
+    next: str
+
+# Create LLM for the supervisor
+llm = ChatOpenAI(model="gpt-4-1106-preview")
+
+# Define the system prompt for the supervisor
+system_prompt = (
+    "You are a supervisor tasked with managing a conversation between the following workers: Researcher, Coder, Viewer. "
+    "Given the following user request, respond with the worker to act next. Each worker will perform a task and respond with their results and status. "
+    "When finished, respond with FINISH. If the request seems like a product review or sentiment analysis, route it to the Viewer."
+)
+
+# Define options
+options = ["FINISH", "Researcher", "Coder", "Viewer"]
+
+# Define the function for routing
+function_def = {
+    "name": "route",
+    "description": "Select the next role.",
+    "parameters": {
+        "title": "routeSchema",
+        "type": "object",
+        "properties": {
+            "next": {
+                "title": "Next",
+                "anyOf": [
+                    {"enum": options},
+                ],
+            }
+        },
+        "required": ["next"],
+    },
+}
+
+# Define the prompt for the supervisor
+prompt = ChatPromptTemplate.from_messages(
+    [
+        ("system", system_prompt),
+        MessagesPlaceholder(variable_name="messages"),
+        (
+            "system",
+            "Given the conversation above, who should act next? Or should we FINISH? Select one of: {options}",
+        ),
+    ]
+).partial(options=str(options), members="Researcher, Coder, Viewer")
+
+# Create the supervisor chain
+supervisor_chain = (
+    prompt
+    | llm.bind_functions(functions=[function_def], function_call="route")
+    | JsonOutputFunctionsParser()
+)
+
+# Define agents
+research_agent = create_agent(llm, [tavily_tool], "You are a web researcher.")
+research_node = functools.partial(agent_node, agent=research_agent, name="Researcher")
+
+code_agent = create_agent(
+    llm,
+    [python_repl_tool],
+    "You may generate safe python code to analyze data and generate charts using matplotlib.",
+)
+code_node = functools.partial(agent_node, agent=code_agent, name="Coder")
+
+viewer_node = functools.partial(viewer_agent)
+
+# Create the workflow
+workflow = StateGraph(AgentState)
+workflow.add_node("Researcher", research_node)
+workflow.add_node("Coder", code_node)
+workflow.add_node("Viewer", viewer_node)
+workflow.add_node("supervisor", supervisor_chain)
+
+# Add edges for each agent to report back to the supervisor
+members = ["Researcher", "Coder", "Viewer"]
+for member in members:
+    workflow.add_edge(member, "supervisor")
+
+# Add conditional edges
+conditional_map = {k: k for k in members}
+conditional_map["FINISH"] = END
+workflow.add_conditional_edges("supervisor", lambda x: x["next"], conditional_map)
+
+# Set the entry point
+workflow.set_entry_point("supervisor")
+
+# Compile the workflow
+graph = workflow.compile()
+
+# Testing the workflow
+for s in graph.stream(
+    {
+        "messages": [
+            HumanMessage(content="write a report of gopal who worked in 3 k technologies")
+        ]
+    }
+):
+    if "__end__" not in s:
+        print(s)
+        print("----")
+
+
+
+