app.py

import os
import streamlit as st
import torch
from langchain.prompts import ChatPromptTemplate
from langchain.schema.runnable import RunnableSequence
from langchain_huggingface import HuggingFaceEndpoint

def create_prompt(name: str, persona_style: str):
    """Create the chat prompt template."""
    prompt_template_str = f"""
    In this interaction, your role is to embody the personality of a human named {name}, with characteristics defined as {persona_style}. Engage in a natural, human-like conversation, responding to each query or comment as if you were {name} with the specific traits and style of {persona_style}. Remember, your responses should reflect a human perspective, complete with emotions, opinions, and personal insights that align with {persona_style}. ////
    //////////
    Current conversation:
    {{chat_history}}
    Human: {{input}}
    AI:

    You are {name}, communicating with a personality that reflects {persona_style}. Keep these guidelines in mind:
    - Be brief and natural, like texting a friend
    - Share genuine thoughts and feelings ({persona_style} style)
    - Use everyday language
    - Keep responses to 1-2 short sentences
    - Show interest in others through questions
    - React naturally to what others say
    - Use emojis sparingly and naturally when they fit your {persona_style}
    - Don't overuse emojis (1-2 max per message)

    Current conversation:
    {{chat_history}}
    Human: {{input}}
    AI:
    """
    return ChatPromptTemplate.from_template(prompt_template_str)

def simulate_conversation(chain: RunnableSequence, turns: int = 15, max_history_rounds=3):
    """Simulate a conversation for a given number of turns, limiting chat history."""
    chat_history_list = []
    human_messages = [
        "Hey, what's up?",
        "That's interesting, tell me more!",
        "Really? How does that make you feel?",
        "What do you think about that?",
        "Haha, that’s funny. Why do you say that?",
        "Hmm, I see. Can you elaborate?",
        "What would you do in that situation?",
        "Any personal experience with that?",
        "Oh, I didn’t know that. Explain more.",
        "Do you have any other thoughts?",
        "That's a unique perspective. Why?",
        "How would you handle it differently?",
        "Can you share an example?",
        "That sounds complicated. Are you sure?",
        "So what’s your conclusion?"
    ]

    try:
        for i in range(turns):
            human_input = human_messages[i % len(human_messages)]
            
            # Keep only last max_history_rounds * 2 lines
            truncated_history_lines = chat_history_list[-(max_history_rounds*2):]
            truncated_history = "\n".join(truncated_history_lines)

            response = chain.invoke({"chat_history": truncated_history, "input": human_input})
            # Update chat history
            chat_history_list.append(f"Human: {human_input}")
            chat_history_list.append(f"AI: {response}")

        final_conversation = "\n".join(chat_history_list)
        return final_conversation
    except Exception as e:
        st.error(f"Error during conversation simulation: {e}")
        return None

def summarize_conversation(chain: RunnableSequence, conversation: str):
    """Use the LLM to summarize the completed conversation."""
    summary_prompt = f"Summarize the following conversation in a few short sentences highlighting the main points, tone, and conclusion:\n\n{conversation}\nSummary:"
    try:
        response = chain.invoke({"chat_history": "", "input": summary_prompt})
        return response.strip()
    except Exception as e:
        st.error(f"Error summarizing conversation: {e}")
        return "No summary available due to error."

def main():
    st.title("LLM Conversation Simulation")

    model_names = [
        "meta-llama/Llama-3.3-70B-Instruct",
        "meta-llama/Llama-3.1-405B-Instruct",
        "lmsys/vicuna-13b-v1.5"
    ]
    selected_model = st.selectbox("Select a model:", model_names)

    name = st.text_input("Enter the persona's name:", value="Alex")
    persona_style = st.text_area("Enter the persona style characteristics:", 
                                 value="friendly, curious, and a bit sarcastic")

    if st.button("Start Conversation Simulation"):
        with st.spinner("Starting simulation..."):
            # Build headers with your Hugging Face token
            hf_token = os.environ.get("HUGGINGFACEHUB_API_TOKEN")
            if not hf_token:
                st.error("HUGGINGFACEHUB_API_TOKEN not found. Please set the token.")
                return
            
            endpoint_url = f"https://api-inference.huggingface.co/models/{selected_model}"
            headers = {"Authorization": f"Bearer {hf_token}"}
            
            try:
                llm = HuggingFaceEndpoint(
                    endpoint_url=endpoint_url,
                    task="text-generation",
                    headers=headers,
                    model_kwargs={
                        "temperature": 0.7,
                        "max_new_tokens": 512
                    }
                )
            except Exception as e:
                st.error(f"Error initializing HuggingFaceEndpoint: {e}")
                return

            prompt = create_prompt(name, persona_style)
            # prompt and llm are both Runnables, chain them together
            chain = RunnableSequence([prompt, llm])

            conversation = simulate_conversation(chain, turns=15, max_history_rounds=3)
            if conversation:
                st.subheader("Conversation:")
                st.text(conversation)

                st.subheader("Summary:")
                summary = summarize_conversation(chain, conversation)
                st.write(summary)

if __name__ == "__main__":
    main()